Method for estimating state of charge of a battery

Abstract

A method for estimating a state of charge of a battery includes the steps of setting an initial state of charge value and an initial Kalman error covariance value, receiving an estimated G parameter value, a current current value, and a current voltage value of the battery, updating a current state of charge value and a current Kalman error covariance value of the battery by inputting the estimated G parameter value, the current current value, and the current voltage value to an extended Kalman filter, and outputting the current state of charge value.

Description

Technical Field

[0001] The disclosure relates to a method for estimating the state of charge (SOC) of a battery in use. Background Technology

[0002] Batteries are readily applicable to electrical devices and possess relatively high energy and power density compared to other energy storage devices. Therefore, they are widely used not only in portable electronic devices but also in electric vehicles (EVs) and hybrid electric vehicles (HEVs) powered by electric drive sources. In particular, when high output is required, battery packs can be used, consisting of multiple individual battery cells connected in series or parallel.

[0003] Battery management is crucial for the energy-efficient and safe use of electrical devices powered by batteries or battery packs, and therefore, accurate measurement and diagnosis of battery status are essential. Currently widely used estimates include State of Charge (SOC), State of Health (SOH), and Power Limit Estimate (PLE).

[0004] According to existing technologies, the State of Charge (SOC) value is represented by the battery's charge level, remaining capacity, etc., and is defined as the percentage of the battery's current capacity relative to its fully charged capacity (or full charge capacity). Methods for estimating SOC include measuring the amount of charge discharged or flowing in using a current sensor and then integrating the charge; using the relationship between open-circuit voltage (OCV) and SOC; and estimating SOC using a battery model.

[0005] The State of Charge (SOC) is estimated by measuring the amount of charge discharged or flowing in using a current sensor. The current SOC is estimated by adding the value obtained by dividing the integrated current value by the full charge capacity to the initial SOC. This method is also known as the Ah counting method or the coulomb counting method and is widely used due to its simplicity; however, it is affected by the accuracy of the current sensor.

[0006] The method of estimating SOC by measuring OCV utilizes a unique OCV-SOC relationship for each battery. It is known that the OCV-SOC relationship changes very little even as the battery deteriorates; therefore, this method is highly reliable. However, to measure OCV, the battery must be kept in a zero-current state for an extended period. Therefore, OCV cannot be measured while the battery is in use, and accurate prediction of SOC is difficult.

[0007] When using battery models, the impact of errors (noise) from current sensors can be minimized, and the State of Charge (SOC) can be estimated in real time during extended periods of idle operation without a battery. Battery models include equivalent circuit models (ECMs), physics-based models, etc. ECMs do not provide insight into what happens within a single battery cell, and the parameters used in ECMs do not actually have physical indications. Physics-based models are more accurate than ECMs, but they suffer from complexity and convergence issues.

[0008] As mentioned above, previous methods for estimating State of Charge (SOC) have limitations, such as the complexity of battery models or the pause in battery use. To prevent overcharging and over-discharging of the battery and to perform cell balancing, accurate SOC estimation is essential. However, Ah-counting methods cannot guarantee accuracy due to current measurement errors. For example, based solely on the assumption that the current sensor has an error of only 0.1A, an SOC estimation error of 0.8Ah occurs when an electric vehicle is used for 8 hours; that is, a weekly SOC estimation error equal to or greater than 5Ah. When the battery capacity is 100Ah, the error reaches 5%.

[0009] For State of Charge (SOC) estimation, not only accuracy is important, but also low computational load and high computational speed are crucial. When the internal state of a battery can be accurately estimated and controlled, the safety and performance of price- and weight-based battery packs can be improved, enabling batteries to be used not only in vehicles but also in transportation (such as aviation) and a wide variety of other fields. Summary of the Invention

[0010] Technical issues

[0011] This disclosure provides a method for accurately estimating the state of charge (SOC) of a battery in real time using the battery's voltage and current values. It also discloses a method for accurately estimating the SOC of a battery in real time using G-parameters that indicate the internal state of the battery. According to the disclosure, the SOC estimation method can be loaded onto a battery management system (BMS) and executed by the BMS, and the SOC of a battery in use can be estimated based on the disclosed SOC estimation method.

[0012] Technical solution

[0013] According to one aspect of the disclosure, a method for evaluating the state of charge (SOC) of a battery includes: setting an initial SOC value and an initial Kalman error covariance value; receiving estimated G-parameter values, a current value, and a current voltage value of the battery; updating the current SOC value and the current Kalman error covariance value of the battery by inputting the estimated G-parameter values, the current current value, and the current voltage value into an extended Kalman filter; and outputting the current SOC value. According to another aspect of the disclosure, the method for estimating the SOC of a battery can be performed by a computing device.

[0014] Publicly disclosed beneficial effects

[0015] Compared to previous methods, the method for estimating the state of charge (SOC) of a battery according to the various disclosed embodiments offers significant improvements in cost, scalability, and adaptability. While previous battery model-based SOC estimation methods were difficult to apply to battery management systems (BMS) due to their complexity, the disclosed SOC estimation method can indeed be implemented on a battery management system (BMS).

[0016] Furthermore, while previous methods required pausing battery use to improve SOC accuracy, the disclosed SOC estimation method allows for real-time estimation of the SOC of a battery in actual use. Additionally, although previous methods limited SOC estimation accuracy due to current sensor errors, the disclosed SOC estimation method minimizes the impact of current sensor errors by employing a filter. Moreover, the disclosed SOC estimation method can be applied not only to individual battery cells or battery packs but also to battery systems in general. Attached Figure Description

[0017] Figure 1 A schematic structural diagram of a battery system according to an embodiment is shown.

[0018] Figure 2 A diagram showing the internal structure of a method for performing state of charge (SOC) estimation according to an embodiment is illustrated.

[0019] Figure 3 A flowchart illustrating the operation of estimating the State of Occurrence (SOC) performed by an extended Kalman filter, according to an embodiment, is shown.

[0020] Figure 4 A flowchart illustrating the operation of calculating G parameter values ​​performed by the GH estimator according to an embodiment is shown.

[0021] Figure 5 The SOC-OCV curve is shown, representing the open-circuit voltage (OCV) relative to the battery's SOC.

[0022] Figure 6 A diagram showing the internal structure of a method for performing SOC estimation according to another embodiment is illustrated.

[0023] Figure 7 A flowchart illustrating the operation of a noise filter according to another embodiment is shown.

[0024] Figure 8It is a graph used to compare the estimated SOC value based on public data with the actual SOC value of the battery.

[0025] Figure 9 It is a graph used to compare the estimated cell voltage based on public data with the actual cell voltage of the battery. Detailed Implementation

[0026] The advantages and features of the disclosure, as well as the methods of implementing it, will become apparent from the embodiments described in detail with reference to the accompanying drawings. However, the disclosure is not limited to the embodiments provided below. Rather, the disclosure can be implemented in a variety of different forms and should be understood to include all modifications, equivalents, or substitutions contained within the concept and scope of the disclosure. The embodiments described below are provided to fully disclose the disclosure so that those skilled in the art can thoroughly understand its scope. In describing the disclosure, well-known techniques in the art will not be described in detail where it is determined that a detailed description of these techniques would unnecessarily obscure the concept of the disclosure.

[0027] The terminology used in this specification is only for describing specific embodiments disclosed and is not intended to limit the disclosure. As used herein, unless the context clearly indicates otherwise, the singular terms “a” and “an” are also intended to include the plural forms. The terms “comprising,” “having,” etc., as used herein should be understood to indicate the presence of the features, quantities, steps, operations, components, parts, or combinations thereof described herein, and should not be construed as preemptively excluding the possibility of the presence or addition of one or more other features, quantities, steps, operations, components, parts, or combinations thereof. Although the terms first, second, etc., may be used herein to describe various components, these terms do not limit the components. These terms are only used to distinguish one component from another.

[0028] The disclosed embodiments are described in detail below with reference to the accompanying drawings. In the description with reference to the drawings, the same reference numerals refer to the same elements, and the same descriptions will not be repeated.

[0029] Figure 1 A schematic structural diagram of a battery system according to an embodiment is shown.

[0030] Reference Figure 1 The battery system 100 may include a battery 110, a voltage measuring unit 120, a current measuring unit 130, a microprocessor 140, and a memory 150.

[0031] Battery 110 is the part that stores electricity and includes a plurality of battery cells 111 electrically connected to each other between a first terminal 101 and a second terminal 102. The battery cells 111 can be connected in series or in parallel, or in a combination of series and parallel connections. The battery cells 111 can have the same capacity and can discharge and charge the same amount of current. However, the internal states of the battery cells 111 can differ from each other. For example, the battery cells 111 can have different internal resistances and previously stored electrical energy. The battery cells 111 can have different G-parameter values ​​and H-parameter values. The battery 110, including the battery cells 111, can also have its own G-parameter and H-parameter values.

[0032] In this disclosure, a method for estimating the state of charge (SOC) of battery 110 is described. However, the method for estimating SOC according to the disclosure can also be applied to the estimation of the SOC of each individual battery cell 111.

[0033] Battery cell 111 may include a rechargeable secondary battery. For example, battery cell 111 may include a nickel-cadmium battery, a lead-acid battery, a nickel metal hydride (NiMH) battery, a lithium-ion battery, a lithium polymer battery, etc. The number of battery cells 111 included in battery 110 can be determined according to the required capacity, output voltage, and output current of battery 110.

[0034] Figure 1 A battery 110 is shown. However, multiple batteries 110 can be connected in parallel and / or in series with each other, and connected to a load and / or charging device via a first terminal 101 and a second terminal 102. Although in Figure 1 Although not shown, battery 110 can be in use while connected to a load and / or charging device. Furthermore, according to the disclosure, the method for estimating SOC can also be applied to the estimation of the SOC of all batteries 110 connected in parallel and / or in series with each other.

[0035] Battery 110 may be a battery pack or battery module including at least one battery cell 111. Battery system 100 may be a system including at least one battery pack or at least one battery module.

[0036] The voltage measurement unit 120 can periodically generate the voltage value of the battery 110 by measuring the voltage between the two electrodes of the battery 110 for each predetermined sampling period Ts. As another example, the voltage measurement unit 120 can periodically generate the voltage value of each battery cell 111 by measuring the voltage of each individual battery cell 111 for each sampling period Ts. When the battery cells 111 are connected in parallel with each other, the voltage measurement unit 120 can measure the voltage of only one battery cell 111 and can determine that the voltage values ​​of all battery cells 111 are the same.

[0037] For example, the sampling period Ts can be one second. However, the sampling period Ts can be set to another time period, such as 0.1 seconds, 0.5 seconds, 2 seconds, 5 seconds, or 10 seconds. The sampling period Ts can be appropriately set according to the electrical system connected to the battery system 100. The currently measured voltage value is called the current voltage value and is denoted as V(t). The voltage value measured before the sampling period Ts is called the previous voltage value and is denoted as V(t-1).

[0038] The current measurement unit 130 can periodically generate the current value of the battery 110 by measuring the current flowing through the battery 110 for each sampling period Ts. As another example, the current measurement unit 130 can periodically generate the current value of each battery cell 111 by measuring the current of each battery cell 111 for each sampling period Ts. When the battery cells 111 are connected in series with each other, the current measurement unit 130 can measure the current of only one battery cell 111 and can determine that the current values ​​of all battery cells 111 are the same.

[0039] The current value measured by the current measuring unit 130 is represented as a positive (+) value in the case of charging current and a negative (-) value in the case of discharging current. The currently measured current value is called the current current value and is represented as I(t), and the current value measured before the sampling period Ts is called the previous current value and is represented as I(t-1). The voltage measuring unit 120 and the current measuring unit 130 can be synchronized with each other and can measure the voltage and current of the battery 110 respectively in the same timing sequence.

[0040] Microprocessor 140 can estimate the state of charge (SOC) of battery 110. Microprocessor 140 can be configured to set an initial SOC value. est (0) and the initial Kalman error covariance value P k (0), receive the estimated G-parameter value (hereinafter referred to as "estimated G-parameter value") from battery 110. est (t), current current value I(t), and current voltage value V(t), by estimating the G parameter value G est The current current value I(t) and the current voltage value V(t) are input to the extended Kalman filter to update the current SOC value of battery 110. est (t) and the current Kalman error covariance value P k (t), and output the current SOC value SOC. est (t).

[0041] Based on the voltage value of battery 110 provided by voltage measurement unit 120 and the current value of battery 110 provided by current measurement unit 130, microprocessor 140 can update the estimated G parameter value G in real time. est (t) and the estimated H parameter values ​​(hereinafter referred to as "estimated H parameter values") H est (t), where the estimated G parameter values ​​are G(t), est (t) and estimated H parameter values ​​H est (t) represent the values ​​of the G parameter and H parameter, respectively, indicating the current state of battery 110. The G parameter represents the sensitivity of the voltage of battery 110 to changes in current, and the H parameter represents the effective potential determined by the local equilibrium potential distribution and resistance distribution in battery 110.

[0042] By using an adaptive filter, the microprocessor 140 can generate estimated G parameter values ​​G in real time based on the voltage and current values ​​of the battery 110. est (t) and estimated H parameter values ​​H est (t). The adaptive filter can be a filter using the recursive least squares (RLS) method or a filter using the weighted least squares (WLS) method. This specification describes in detail how the microprocessor 140 generates estimated G parameter values ​​G of the battery 110 based on the voltage and current values ​​of the battery 110 using an RLS filter. est (t) and estimated H parameter values ​​H est Examples of (t) are provided. However, the disclosure is not limited thereto.

[0043] Microprocessor 140 can input the estimated G parameter value G of battery 110 into the extended Kalman filter. est The current SOC value of battery 110 is generated in real time using the current current value I(t) and the current voltage value V(t). est (t), where the estimated G parameter values ​​are G(t), est (t) is generated using an adaptive filter. The microprocessor 140 can use coefficient data stored in the memory 150. The coefficient data can be generated based on a predetermined open-circuit voltage (OCV) - state of charge (SOC) relationship with respect to the battery 110.

[0044] The microprocessor 140 uses only simple operations (such as the four basic arithmetic operations) to generate the current SOC value of the battery 110 in real time. est(t), therefore, the microprocessor 140 may be included in the battery management system (BMS) installed in the battery system 100 or battery pack. As another example, the method for estimating the State of Charge (SOC) according to this embodiment may be executed by the microcontroller or electronic control unit (ECU) of the BMS of the electric vehicle. As another example, the method for estimating the SOC according to this embodiment may be executed by the integrated controller of the energy storage system. As another example, the method for estimating the SOC according to this embodiment may be executed by the processor of a server connected to the battery system or energy storage system for communication.

[0045] According to this embodiment, the memory 150 can store the instructions and data required by the microprocessor 140 to execute the method for estimating the state of charge (SOC). According to this embodiment, based on the voltage and current values ​​of the battery 110 generated for each sampling period Ts, an estimated G parameter value G for the battery 110 can be generated. est (t), and based on the estimated G parameter values ​​G est The current current value I(t) and the current voltage value V(t) can be used to generate the current state of charge (SOC) value. est (t). Therefore, the current voltage value, the current current value, the previous current value, etc., can be stored in memory 150, while other voltage and current data may not be stored in memory 150. Memory 150 does not need to store a large number of instructions and data; therefore, memory 150 can be implemented as a small memory. For example, memory 150 can be implemented as the memory in microprocessor 140.

[0046] The disclosure provides a method for estimating the State of Charge (SOC) of a battery using G-parameters and H-parameters representing the current state of the battery. According to the disclosure, the method for estimating SOC can be implemented relatively simply in a BMS and can achieve high accuracy without additional operating conditions.

[0047] The G-parameter is a state quantity representing the sensitivity of the voltage at the terminals of a battery cell in use to changes in the current applied to the cell, and is expressed in units of resistance. The H-parameter is the effective potential determined by the local equilibrium potential distribution and resistance distribution within the battery cell in use. The G-parameters and H-parameters of a battery cell can be quantified using theoretical models through explicit correlation formulas relating battery material properties and design variables. The G-parameters and H-parameters of a battery cell are described below.

[0048] Regarding a single battery cell, we can assume that the voltage V and current I have a relationship such as V = f(I; x, p). Here, x is a physical quantity representing the internal state of the battery cell, and p is a parameter.

[0049] The function f is a nonlinear implicit function, and when the function f can be divided into a rapidly changing quantity g and a slowly changing quantity h, the voltage V and the current I can be expressed as V = g(I; x, p) + h(I; x, p).

[0050] Assume there exists a function that changes slowly with respect to the current I. Then the voltage V and current I can be expressed as V = G(I; x, p) × I + H(I; x, p). Here, and It has very small values. In other words, when the described assumptions are satisfied, G and H are functions that change slowly with respect to the current I. Therefore, the function f representing the nonlinear relationship between voltage V and current I can be represented by the quasi-linear relationship as described above.

[0051] Here, G is called the G parameter, and H is called the H parameter. When the current I is the charging and discharging current, and Ueq is the equilibrium potential of a single battery cell, the discharge overvoltage can be expressed by Ueq-V=-G×I+(Ueq-H) using the G parameter G and the H parameter H.

[0052] Here, -G×I is the overvoltage caused by current leakage through the battery terminals, and includes the dynamic polarization of the reaction and the electronic and ionic resistivity polarization. (Ueq-H) is the overvoltage caused by the deviation of the local thermodynamic equilibrium state in the battery from the equilibrium state of the general system. That is, (Ueq-H) represents the low efficiency caused by thermodynamic non-uniformity in the battery, and when the internal system of the battery reaches thermodynamic equilibrium, the H parameter H becomes equal to the equilibrium potential Ueq.

[0053] According to the disclosed embodiments, the method for estimating SOC can directly extract the battery's G parameter G from the battery's voltage and current values, and the battery's SOC can be estimated using the G parameter G.

[0054] Figure 2 A diagram showing the internal structure of a method for performing SOC estimation according to an embodiment is illustrated.

[0055] Reference Figure 2 as well as Figure 1The microprocessor 140 may include a GH estimator 142 and an extended Kalman filter 144. The memory 150 may store a lookup table 152 generated based on a predetermined OCV-SOC relationship with respect to the battery 110. The lookup table 152 may store coefficient data corresponding to the SOC values. Since the OCV-SOC relationship is non-linear, the coefficient data corresponding to the OCV-SOC relationship can be stored in the lookup table 152, and the extended Kalman filter 144 can use the coefficient data stored in the lookup table 152 to estimate the SOC of the battery 110. The coefficient data may include first coefficient data or may include second and third coefficient data.

[0056] Figure 1 The voltage measurement unit 120 generates a voltage value V of the battery 110 and provides the voltage value V to the GH estimator 142 and the extended Kalman filter 144. The voltage value V includes the current voltage value V(t) and the previous voltage value V(t-1). After a sampling period Ts, the current voltage value V(t) becomes the previous voltage value V(t-1), and the new voltage value becomes the current voltage value V(t).

[0057] Figure 1 The current measurement unit 130 generates a current value I of the battery 110 and provides the current value I to the GH estimator 142 and the extended Kalman filter 144. The current value I includes the current value I(t) and the previous current value I(t-1). After a sampling period Ts, the current value I(t) becomes the previous current value I(t-1), and the new current value becomes the current value I(t).

[0058] GH estimator 142 generates estimated G parameter values ​​G of battery 110 based on voltage value V and current value I using an adaptive filter. est The adaptive filter can be, for example, a filter using the RLS method. The estimated G parameter values ​​G for battery 110. est This is the value of the G parameter, which represents the sensitivity of the battery 110's voltage to changes in current. For example, the GH estimator 142 generates an estimated H parameter value H for the battery 110 based on the voltage value V and the current value I using an adaptive filter. est The estimated H parameter value H of battery 110. est This is the numerical value of the H parameter, which represents the effective potential determined by the local equilibrium potential distribution and resistance distribution in battery 110. See below for reference. Figure 4 The operation of GH estimator 142 is described in further detail.

[0059] The extended Kalman filter 144 can receive the estimated G parameter values ​​G generated by the GH estimator 142. est The voltage value V and the current value I, and based on the estimated G parameter value Gest The voltage value V and the current value I generate the SOC value. est The extended Kalman filter 144 can use coefficient data stored in lookup table 152. The extended Kalman filter 144 can update the SOC value. est And the Kalman error covariance value, so that whenever a new input is input, the estimated G parameter value G is obtained. est Real-time output of SOC value based on voltage value V and current value I. est .

[0060] According to the embodiment, based on the current current value I(t), the first-order estimated SOC value and the first-order estimated Kalman error covariance value can be calculated. This is based on the coefficient data stored in lookup table 152, the current current value I(t), and the estimated G parameter value G. est This allows for the calculation of the estimated voltage value and the current Kalman gain. Based on the first-order estimated SOC value, the current Kalman gain value, the current voltage value V(t), and the estimated voltage value, the current SOC value can be updated. Furthermore, based on the first-order estimated Kalman error covariance value, the current Kalman gain value, the coefficient data, and the estimated G parameter value G... est This can update the current Kalman error covariance value.

[0061] The following reference Figure 3 The operation of the extended Kalman filter 144 is described in further detail.

[0062] Figure 3 A flowchart illustrating the operation of estimating the State of Occurrence (SOC) performed by an extended Kalman filter according to an embodiment is shown.

[0063] Reference Figures 1 to 3 The operation of the extended Kalman filter 144 is performed by the microprocessor 140.

[0064] In operation S10, the microprocessor 140 can receive the initial SOC value SOC. est (0) and the initial Kalman error covariance value P k (0), and set the initial SOC value SOC. est (0) and the initial Kalman error covariance value P k (0). Initial SOC value. est (0) and the initial Kalman error covariance value P k (0) are respectively used as the previous SOC value SOC. est (t-1) and the previous Kalman error covariance value P k (t-1). Users can estimate the SOC of battery 110 or input any value as the initial SOC value. est (0) to input the initial SOC value. est(0). Users can input any value as the initial Kalman error covariance value P. k (0).

[0065] In operation S20, the extended Kalman filter 144 of the microprocessor 140 can receive the estimated G parameter value G for each sampling period Ts. est The microprocessor 140 can estimate the G parameter values ​​G(t), the current current value I(t), and the current voltage value V(t). est The current value V(t) is input to the extended Kalman filter 144. The microprocessor 140 can receive the current voltage value V(t) from the voltage measurement unit 120 and the current current value I(t) from the current measurement unit 130.

[0066] In operation S30, the microprocessor 140 can base its operation on the previous SOC value. est (t-1), current value I(t), sampling period Ts, and maximum capacity Q of battery 110 max To calculate the first-order estimated SOC value. est -(t). The microprocessor 140 can store information about the sampling period Ts and the maximum capacity Q of the battery 110. max Information. Previous SOC value. est (t-1) and the current SOC value before the sampling period Ts. est (t) corresponds. This can be based on SOC. est -(t)=SOC est (t-1)+I(t)×Ts / Q max To calculate the first-order estimated SOC value. est -(t).

[0067] In operation S40, the microprocessor 140 can base its operation on the previous Kalman error covariance value P. k (t-1), sampling period Ts, maximum capacity Q max and processor noise σ w To calculate the first-order estimated Kalman error covariance value P k -(t). Processor noise σ w This can be set by the user according to the system specifications, for example, set to 10- 6 With 10- 1 The values ​​between [a certain range]. Processor noise σ w It can be set to, for example, 0.0002. The previous Kalman error covariance value P k (t-1) and the current Kalman error covariance value P before the sampling period Ts. k (t) corresponds. This can be based on P. k -(t)=Pk (t-1)+(Ts / Q max ) 2 ×σ w To calculate the first-order estimated Kalman error covariance value P k -(t).

[0068] In operation S50, the microprocessor 140 can base its operation on coefficient data, the current value I(t), and the estimated G parameter value G. est (t) is used to calculate the estimated voltage value V. est (t).

[0069] According to an embodiment, the microprocessor 140 can extract the first-order estimated SOC value (SOC) calculated in operation S30 from the coefficient data stored in lookup table 152. est The first coefficient value C(t) corresponding to (t).

[0070] Coefficient data can be generated based on a predetermined OCV-SOC relationship for battery 110. For example, Figure 5 The curve representing OCV relative to SOC is shown. (Example) Figure 5 As shown, OCV is non-linear with respect to SOC. Lookup table 152 stores the data of OCV relative to SOC, which is related to... Figure 5 The OCV curve shown corresponds to the SOC curve. Lookup table 152 must be stored in memory 150; therefore, lookup table 152 stores some OCV data values ​​corresponding to some SOC data values. For example, some SOC data values ​​can be, for example, 0.01, 0.05, 0.1, 0.2, 0.5, 0.8, 0.9, 0.95, and 0.99, and lookup table 152 can store the OCV data values ​​corresponding to the SOC data values ​​0.01, 0.05, 0.1, 0.2, 0.5, 0.8, 0.9, 0.95, and 0.99 as coefficient data, respectively. As another example, some SOC data values ​​can be, for example, multiples of 0.05 or 0.1.

[0071] It is possible to estimate the SOC value based on a value close to the first order. est -(t) SOC data value N ε [SOC est -(t)] and the SOC data value N ε [SOC est -(t)] corresponds to the OCV data value OCV(N) ε [SOC est The first coefficient value C(t) can be determined based on C(t) = OCV(N). ε [SOC est -(t)]) / N ε [SOCest -(t)] is used to determine the first coefficient value C(t). For example, when the first-order estimated SOC value is SOC est When -(t) is 0.11, the SOC data value N is close to 0.11. ε [SOC est -(t)] can be 0.1. When the OCV data value corresponding to the SOC data value of 0.1 is 3, the first coefficient value C(t) can be 30 corresponding to 3 / 0.1, and these values ​​can be stored as coefficient data in memory 150. SOC data value N ε [SOC est [-(t)] can be the closest first-order estimate of SOC value in the coefficient data. est -(t) SOC data value.

[0072] Microprocessor 140 can estimate the SOC value based on the first coefficient value C(t) and the first-order estimated SOC value. est -(t), Estimated G parameter values ​​G est The estimated voltage value V is calculated using I(t) and the current value I(t). est (t). Can be based on V est (t)=C(t)×SOC est -(t)+G est The estimated voltage value V is calculated by multiplying I(t) by I(t). est (t). C(t)×SOC est -(t) is the same as the first-order estimated SOC value. est -(t) corresponds to the OCV value, and G est (t)×I(t) corresponds to the voltage drop caused by the resistive component of battery 110.

[0073] According to another embodiment, the microprocessor 140 can extract the first-order estimated SOC value (SOC) calculated in operation S30 from the coefficient data stored in lookup table 152. est -(t) corresponds to the second coefficient value C1(t) and the third coefficient value E(t). Looking up Table 152, the first-order estimated SOC value SOC can be obtained. est -(t) stores coefficient data including the second coefficient value C1(t) and the third coefficient value E(t).

[0074] The coefficient data can be determined based on the predetermined OCV-SOC relationship for battery 110. (Refer to...) Figure 5 The SOC-OCV curve represents the OCV relative to the SOC. The second coefficient value C1(t) and the third coefficient value E(t) can be determined as the contact coefficients. Figure 5 The slope and OCV intercept of the linear function of the SOC-OCV curve at a point close to the first-order estimated SOC value.est -(t) SOC data value N ε [SOC est -(t)] corresponds. For example, when the SOC data value N ε [SOC est -(t)] is Figure 5 SOC a At that time, the second coefficient value C1(t) and the third coefficient value E(t) can be determined as linear functions TL of the points contacting the SOC-OCV curve. a slope C a and OCV intercept E a This point is related to SOC a Correspondingly. As another example, when the SOC data value N ε [SOC est -(t)] is Figure 5 SOC b At that time, the second coefficient value C1(t) and the third coefficient value E(t) can be determined as linear functions TL of the points contacting the SOC-OCV curve. b slope C b and OCV intercept E b This point is related to SOC b correspond.

[0075] Microprocessor 140 can base its calculations on the second coefficient value C1(t) and the first-order estimated SOC value SOC. est -(t), Estimated G parameter values ​​G est The estimated voltage value V is calculated using the current value I(t) and the third coefficient value E(t). est (t). Can be based on V est (t)=C1(t)×SOC est -(t)+G est The estimated voltage value Vest(t) is calculated using C1(t)×I(t)+E(t). est -(t)+E(t) is the same as the first-order estimated SOC value SOC. est -(t) corresponds to the OCV value, and G est (t)×I(t) corresponds to the voltage drop caused by the resistive component of battery 110.

[0076] In operation S60, microprocessor 140 can estimate the Kalman error covariance value P based on coefficient data and a first-order estimate. k -(t) and estimated G parameter values ​​G est (t) is used to calculate the current Kalman gain value L. k (t).

[0077] Microprocessor 140 can extract the first-order estimated SOC value from the coefficient data stored in lookup table 152. est -(t) corresponds to the first coefficient value C(t). According to an embodiment, the first coefficient value C(t) may be the same as the first coefficient value C(t) extracted in operation S50, or according to another embodiment, it may be the same as the second coefficient value C1(t) extracted in operation S50.

[0078] Microprocessor 140 can estimate the Kalman error covariance value P based on the first coefficient value C(t) and the first-order estimate. k -(t), Estimated G parameter values ​​G est (t) and the measured noise (hereinafter referred to as "measurement noise") σ v To calculate the current Kalman gain value L k (t). Measurement of noise σ v The user can set it to, for example, 10- according to the system specifications (e.g., voltage measurement unit 120 and current measurement unit 130). 1 With 10- 3 The value between [variable values]. Measurement of noise σ. v It can be set to, for example, 0.01. Current Kalman gain value L k (t) can be based on L k (t)=C(t)×P k -(t) / [C(t) 2 ×P k -(t)+G est (t) 2 ×σ v To calculate.

[0079] In operation S70, the microprocessor 140 can base its operation on the first-order estimated SOC value calculated in operation S30. est -(t), the current Kalman gain value L calculated in operation S60. k (t), the current voltage value V(t) received in operation S20, and the estimated voltage value V calculated in operation S50. est (t) is used to update and output the current SOC value. est (t). Current SOC value. est (t) can be based on SOC est (t)=SOC est -(t)+L k (t)×(V(t)-V est The SOC value can be calculated using (t). The current SOC value calculated as described above can be output. est (t) is the estimated SOC value of battery 110. Since the current SOC value is SOC... est(t) must be updated and output in real time after the next sampling period Ts, therefore the current Kalman error covariance value P is updated in operation S80. k (t).

[0080] In operation S80, microprocessor 140 can estimate the Kalman error covariance value P based on the first-order estimate. k -(t), Current Kalman gain value L k (t), coefficient data and estimated G parameter values ​​G est (t) is used to update the current Kalman error covariance value P. k (t).

[0081] Microprocessor 140 can extract the first-order estimated SOC value from the coefficient data stored in lookup table 152. est -(t) corresponds to the first coefficient value C(t). According to an embodiment, the first coefficient value C(t) may be the same as the first coefficient value C(t) extracted in operation S50, or according to another embodiment, it may be the same as the second coefficient value C1(t) extracted in operation S50.

[0082] Microprocessor 140 can base its calculation on the first-order estimated Kalman error covariance value P calculated in operation S40. k -(t), the current Kalman gain value L calculated in operation S60. k (t), the first coefficient value C(t), and the estimated G parameter value G received in operation S20. est (t) and the measured noise σ v To calculate the current Kalman error covariance value P k (t). Current Kalman error covariance value P k (t) can be based on P k (t)=P k -(t)-L k (t) 2 ×[C(t) 2 ×P k -(t)+G est (t) 2 ×σ v To calculate.

[0083] After the sampling period Ts in operation S90, the microprocessor 140 can proceed to operation S20, and can receive the new estimated G parameter value G in operation S20. est (t), the new current value I(t), and the new current value V(t). Previous estimated G parameter values ​​G est (t) becomes the previous G parameter value G est(t-1), the previous current value I(t) becomes the previous current value I(t-1), and the previous current voltage value V(t) becomes the previous voltage value V(t-1). Based on the newly received estimated G parameter value G... est Repeating operations S20 to S80 with the current current value I(t) and the current voltage value V(t), a new current SOC value can be calculated and output in real time. est (t).

[0084] Figure 4 A flowchart is shown illustrating the operation of calculating the G parameter values ​​performed by the GH estimator according to an embodiment.

[0085] The operation of the GH estimator 142 is executed via the microprocessor 140. The microprocessor 140 can estimate the G-parameter and H-parameter values ​​of the battery 110 by using an adaptive filter.

[0086] Reference Figure 1 , Figure 2 and Figure 4 When the microprocessor 140 uses the RLS filter, in operation S110, the microprocessor 140 can set the initial state vector value Θ. est (0) = [G est (0); H est [0] and the initial covariance matrix value P(0) = [P1(0); P2(0)]. Initial state vector value Θ est (0) and the initial covariance matrix value P(0) are used as the previous state vector value Θ. est (t-1) and the previous covariance matrix value P(t-1). Current state vector value Θ est The covariance matrix P(t) and the current covariance matrix value P(t) will converge later, therefore, the user can input any value as the initial state vector value Θ. est (0) and the initial covariance matrix value P(0). For example, the initial state vector value Θ est (0) can be set to Θ est (0) = [G est (0); H est [P1(0)] = [1; 1], and the initial covariance matrix value P(0) can be set as P(0) = [P1(0); P2(0)] = [1; 1]. In this example, the initial state vector value Θ est Both (0) and the initial covariance matrix value P(0) are initialized to 1. However, this is only an example, and the initial state vector value Θ... est (0) and the initial covariance matrix value P(0) can be initialized to other values.

[0087] In operation S120, the microprocessor 140 can receive the current voltage value V(t) and the current current value I(t) of the battery 110. The voltage measurement unit 120 and the current measurement unit 130 can sense the voltage and current of the battery 110 for each sampling period Ts, and can provide the voltage value and current value to the microprocessor 140 for each sampling period Ts. The microprocessor 140 can determine the current voltage value or the most recently received voltage value as the current voltage value V(t), and can determine the current current value or the most recently received current value as the current current value I(t). The current voltage value V(t) and the current current value I(t) of the battery 110 received before the sampling period Ts become the previous voltage value V(t-1) and the previous current value I(t-1), respectively.

[0088] The microprocessor 140 can update the current state vector value Θ based on the current voltage value V(t) and the current current value I(t) received for each sampling period Ts. est (t)=[G est (t); H est [(t)] and the current covariance matrix value P(t) = [P1(t); P2(t)]. The current state vector value Θ of battery 110. est (t) Estimated G parameter value G from battery 110 est (t) and estimated H parameter values ​​H est (t) constitutes and is defined as Θ est (t)=[G est (t); H est The current covariance matrix value P(t) is composed of the first covariance matrix value P1(t) and the second covariance matrix value P2(t), and is defined as P(t) = [P1(t); P2(t)].

[0089] For each sampling period Ts, the current voltage value V(t) and current value I(t) of battery 110 are received, and the current state vector value Θ is updated recursively for each sampling period Ts. est The microprocessor 140 can update the current state vector value Θ for each sampling period Ts based on the current voltage value V(t) and the current current value I(t) received for each sampling period Ts using the RLS method. est (t). When updating the current state vector value Θ est When (t), determine the estimated value of the G parameter G. est (t) and estimated H parameter values ​​H est (t).

[0090] In operation S130, the microprocessor 140 calculates the current gain matrix value L(t) based on the first forgetting factor λ1, the second forgetting factor λ2, the current value I(t), and the previous covariance matrix value P(t-1).

[0091] The current gain matrix value L(t) is used to calculate the current state vector value Θ. est The gain matrix value L(t) is composed of the first gain matrix value L1(t) and the second gain matrix value L2(t), and can be calculated as follows.

[0092]

[0093] Here, λ1 is the first forgetting factor and is related to the G parameter. λ2 is the second forgetting factor and is related to the H parameter. The value of the G parameter is relative to the estimated G parameter value. est (t) and estimated H parameter values ​​H est The calculation of (t) uses the first forgetting factor λ1 and the second forgetting factor λ2 to represent the previous voltage value and the previous current value, respectively, relative to the current estimated G parameter value G. est (t) and the current estimated H parameter value H est The influence of (t) on the value of the first forgetting factor λ1 and the second forgetting factor λ2 is as close to 1 as the value of the estimated G parameter G increases. est (t) and estimated H parameter values ​​H est The longer the influence of (t) lasts, and the closer the first forgetting factor λ1 and the second forgetting factors λ2 are to 0, the better the estimation of the G parameter value G becomes. est (t) and estimated H parameter values ​​H est The shorter the influence of (t), the better.

[0094] For example, the first forgetting factor λ1 and the second forgetting factor λ2 can be greater than or equal to 0.9 and less than or equal to 1. As another example, the first forgetting factor λ1 can be set to a value greater than or equal to the value of the second forgetting factor λ2. For example, the first forgetting factor λ1 can be set to 0.99999, and the second forgetting factor λ2 can be set to 0.95. These settings can vary depending on the characteristics of the battery 110.

[0095] The inventors disclosed that, in experiments performed on specific battery cells, highly reliable results were obtained when the first forgetting factor λ1 and the second forgetting factor λ2 were 0.99999 and 0.95, respectively. However, the above values ​​are examples, and different values ​​can be set according to the characteristics of the battery cell 111. For example, the first forgetting factor λ1 can be set to 0.9999, and the second forgetting factor λ2 can be set to 0.98.

[0096] As another example, both the first forgetting factor λ1 and the second forgetting factor λ2 can be set to 1. In this case, it can be considered that the first forgetting factor λ1 and the second forgetting factor λ2 are not applied.

[0097] In operation S140, the microprocessor 140 calculates the current covariance matrix value P(t) based on the first forgetting factor λ1, the second forgetting factor λ2, the current current value I(t), the previous covariance matrix value P(t-1), and the current gain matrix value L(t).

[0098] When calculating the current gain matrix value L(t) after a sampling period Ts, the current covariance matrix value P(t) is used as the previous covariance matrix value P(t-1). The current covariance matrix value P(t) can be calculated as follows.

[0099]

[0100] In operation S150, the microprocessor 140 is based on the previous state vector value Θ. est The current state vector value Θ is calculated using (t-1), the current gain matrix value L(t), the current voltage value V(t), and the current current value I(t). est (t).

[0101] Microprocessor 140 can base its decisions on the current value I(t) and the previous state vector value Θ. est (t-1)=[G est (t-1); H est (t-1)] is used to calculate the estimated voltage value V of battery 110. est (t). For example, it can be based on the current current value I(t) and the previous G parameter value G. est (t-1) and the previous H parameter value H est (t-1) according to V est (t)=G est (t-1)×I(t)+H est (t-1) is used to calculate the estimated voltage value V. est (t).

[0102] Microprocessor 140 can base its calculations on the current voltage value V(t) and the estimated voltage value Vt. est (t) According to e(t)=V(t)-V est The voltage error e(t) is calculated using (t).

[0103] Microprocessor 140 can base its decisions on the previous state vector value Θ. est The current state vector value Θ is calculated using (t-1), the current gain matrix value L(t), and the voltage error e(t). est (t). For example, it can be based on Θ est(t)=Θ est The current state vector value Θ is calculated using (t-1)+L(t)×e(t). est (t). When calculating the current state vector value Θ est When (t), determine the estimated value of the G parameter G. est (t) and estimated H parameter values ​​H est (t).

[0104] In operation S160, microprocessor 140 can use the estimated G parameter value G determined in operation S150. est The values ​​of I(t) and the current current I(t) are provided to the extended Kalman filter 144.

[0105] In operation S170, microprocessor 140 can repeatedly execute operations S120 to S150 for each sampling period Ts.

[0106] The recursive representation of the current state vector value Θ can be derived as follows: est Formula Θ of (t) est (t)=Θ est (t-1)+L(t)×e(t).

[0107] First, the loss function ε, applying the first forgetting factor λ1 and the second forgetting factor λ2, is defined as follows.

[0108]

[0109] Here, V(i) is the i-th voltage value, and I(i) is the i-th current value. V(t) and I(t) are the current voltage value and the current current value, respectively, and V(t-1) and I(t-1) are the previous voltage value and the previous current value, respectively.

[0110] G(i) and H(i) are the i-th actual G parameter value and the i-th actual H parameter value, respectively, and G est (t) and H est (t) represents the current estimated G parameter value and the current estimated H parameter value, respectively.

[0111] When the loss function ε is relative to G est (t) and H est When the derivative of each of the terms in (t) is 0, the loss function ε is relative to G. est (t) and H est (t) becomes minimized.

[0112] The loss function ε is calculated as follows with respect to its derivative to produce Gest(t) with a result of 0.

[0113]

[0114] To correct the above formula, G est (t) is as follows.

[0115]

[0116] The loss function ε is calculated as follows with respect to its derivative to produce a result of 0 for H. est (t).

[0117]

[0118] To correct the above formula, H est (t) is as follows.

[0119]

[0120] For real-time estimation, the current state vector value Θ is used as follows: est (t) recursively modify the G obtained above est (t) and H est (t).

[0121] Θ est (t)=[G est (t); H est [(t)]=Θ est (t-1)+L(t)×[V(t)-G est (t-1)×I(t)-H est (t-1)]

[0122] It can be based on V est (t)=G est (t-1)×I(t)+H est (t-1) is used to calculate the estimated voltage value V. est (t), and based on e(t) = V(t) - V est Let (t) define the voltage error e(t), therefore, the current state vector value Θ est (t) can be represented as described above.

[0123] Θ est (t)=[G est (t); H est [(t)]=Θ est (t-1)+L(t)×e(t)

[0124] Here, as described above, the current gain matrix value L(t) and the current covariance matrix value P(t) are calculated as follows.

[0125]

[0126]

[0127] According to this embodiment, because a recursive method is used to calculate the G parameter value, the memory 150 stores the current voltage value V(t), the current current value I(t), and the current state vector value Θ. est (t), the current covariance matrix value P(t), the first forgetting factor λ1, and the second forgetting factor λ2. The calculation is very simple and can be performed using only a small memory 150 with a size of only a few kB. Because the current state vector value Θ est The values ​​of G(t) and the current covariance matrix P(t) are updated periodically, so the changes in the voltage and current of battery 110 can be reflected in the estimated G parameter value G in real time. est (t) and estimated H parameter values ​​H est (t)

[0128] Figure 6 A diagram showing the internal structure of a method for performing SOC estimation according to another embodiment is illustrated.

[0129] Reference Figure 6 In addition to the GH estimator 142 and the extended Kalman filter 144, the microprocessor 140 may also include a noise filter 146. The GH estimator 142 and the extended Kalman filter 144 have been described as described above, and therefore their descriptions are not repeated.

[0130] According to this embodiment, the sensed voltage value Vsen corresponding to the voltage sensed by the voltage measurement unit 120 can be input to the noise filter 146, and the noise filter 146 can output a voltage value V corresponding to the sensed voltage value Vsen. The voltage value V includes a previous voltage value V(t-1) and a current voltage value V(t). The sensed current value Isen corresponding to the current sensed by the current measurement unit 130 can be input to the noise filter 146, and the noise filter 146 can output a current value I corresponding to the sensed current value Isen. The current value I includes a previous current value I(t-1) and a current value I(t).

[0131] Noise filter 146 can remove noise from the sensed voltage value Vsen and the sensed current value Isen, and can output the noise-removed voltage value V and current value I. Noise filter 146 can be, for example, a low-pass filter. Noise filter 146 can be a moving average filter. Noise filter 146 can be an infinite impulse response (IIR) filter or a finite impulse response (FIR) filter.

[0132] The voltage value V and current value I output from the noise filter 146 can be input to the GH estimator 142. The sensed voltage value Vsen input to the noise filter 146 and the current value I output from the noise filter 146 can be input to the extended Kalman filter 144. The sensed voltage value Vsen and current value I input to the extended Kalman filter 144 correspond to the current voltage value V(t) and the current current value I(t), respectively.

[0133] Figure 7 A flowchart illustrating the operation of a noise filter according to another embodiment is shown.

[0134] Reference Figure 7 In operation S210, a sensed voltage value Vsen corresponding to the voltage of the battery 110 (sensed by the voltage measuring unit 120) is generated, and a sensed current value Isen corresponding to the current of the battery 110 (sensed by the current measuring unit 130) is generated.

[0135] In operation S220, each of the sensed voltage value Vsen and sensed current value Isen is input to the noise filter 146, and a noise-filtered voltage value V and a noise-filtered current value I are generated.

[0136] In operation S230, the voltage value V and the current value I are transmitted to the GH estimator 142, and the current value I and the sensed voltage value Vsen are transmitted to the extended Kalman filter 144. The sensed voltage value Vsen and the current value I input to the extended Kalman filter 144 correspond to the current voltage value V(t) and the current current value I(t), respectively.

[0137] Figure 8 This is a graph comparing the estimated SOC value with the actual (precise) SOC value of the battery according to the embodiment. Figure 9 It is a graph comparing the estimated cell voltage with the actual cell voltage of the battery according to the embodiment.

[0138] like Figure 8 As shown, although the initial SOC value is set to 0.3, the estimated SOC value according to the embodiment follows the actual SOC value after approximately 400 seconds. The gap between the estimated SOC value and the actual SOC value decreases over time. The disclosed method uses simple calculations, therefore, it can be driven by processors with low specifications. Furthermore, the battery's SOC can be estimated with high accuracy according to this method.

[0139] like Figure 9 As shown, the system identifies the actual individual cell voltage based on the publicly available estimated voltage. It is recognized that because the publicly available estimation algorithm is accurate, this method can be applied to a wide variety of products.

[0140] The disclosed concepts should not be limited to the embodiments described above. Furthermore, all scopes equivalent to or modified from the claims, other than those described, should be considered as disclosed concepts.

Claims

1. A method for estimating the state of charge of a battery, the method comprising the following steps: Set the initial state of charge value and the initial Kalman error covariance value; The voltage and current of the battery are sensed for each predetermined sampling period Ts, and the voltage and current values ​​of the battery are generated periodically. An estimated G-parameter value is generated from the voltage value and the current value using an adaptive filter, wherein the estimated G-parameter value is a numerical value of the G-parameter representing the sensitivity of the battery's voltage to changes in current. The estimated G-parameter value, the current current value among the periodically generated current values, and the current voltage value among the periodically generated voltage values ​​are input into the extended Kalman filter. The extended Kalman filter is used to update the current state of charge and the current Kalman error covariance of the battery; and Output the current state of charge value. The steps of updating the current state of charge value and the current Kalman error covariance value of the battery include: Calculate the first-order estimated state of charge value and the first-order estimated Kalman error covariance value based on the current current value; Receive coefficient data generated from the predetermined open-circuit voltage-state-of-charge relationship of the battery; The estimated voltage value and the current Kalman gain value are calculated based on the coefficient data, the current current value, and the estimated G-parameter value; and The current state of charge (SPO) value is updated based on the first-order estimated SPO, the current Kalman gain, the current voltage value, and the estimated voltage value.

2. The method according to claim 1, wherein, The steps of updating the current state of charge value and the current Kalman error covariance value of the battery further include: The current Kalman error covariance is updated based on the first-order estimated Kalman error covariance, the current Kalman gain, the coefficient data, and the estimated G-parameter value.

3. The method according to claim 1, wherein, By using the previous state of charge (SOC) value est (t-1), the current value I(t), the sampling period Ts, and the maximum capacity Q of the battery. max Based on SOC est - (t)=SOC est (t-1)+I(t)×Ts / Q max To calculate the first-order estimated state of charge (SOC) value est - (t).

4. The method according to claim 1, wherein, By using the previous Kalman error covariance value P k (t-1), sampling period Ts, maximum capacity Q of the battery max and processor noise σ w Based on P k - (t)=P k (t-1)+(Ts / Q max ) 2 ×σ w To calculate the first-order estimated Kalman error covariance value P k - (t).

5. The method according to claim 1, wherein, The steps for calculating the estimated voltage value include: Extract the first-order estimated state of charge (SOC) value from the coefficient data. est - The first coefficient value C(t) corresponding to (t); and By using the first coefficient value C(t) and the first-order estimated state of charge value SOC est - (t), the estimated G parameter value G est (t) and the current current value I(t), based on V est (t)=C(t)×SOC est - (t)+G est The estimated voltage value V is calculated by using (t)×I(t). est (t).

6. The method according to claim 5, wherein, By using a value close to the first-order estimated state of charge (SOC) est - The state-of-charge data value N of (t) ε [SOC est - (t)] and the state of charge data value N ε [SOC est - The corresponding open-circuit voltage data value OCV(N)] is [(t)]. ε [SOC est - (t)]), based on C(t)=OCV(N ε [SOC est - (t)]) / N ε [SOC est - The first coefficient value C(t) is determined by [(t)].

7. The method according to claim 1, wherein, The steps for calculating the estimated voltage value include: Extract the first-order estimated state of charge (SOC) value from the coefficient data. est - The second coefficient value C1(t) and the third coefficient value E(t) corresponding to (t); and By using the second coefficient value C1(t), the first-order estimated state of charge value SOC est - (t), the estimated G parameter value G est (t), the current current value I(t), and the third coefficient value E(t), based on V est (t)=C1(t)×SOC est - (t)+G est The estimated voltage value V is calculated by using (t)×I(t)+E(t). est (t).

8. The method according to claim 7, wherein, The second coefficient value C1(t) and the third coefficient value E(t) are determined as the slope and open-circuit voltage intercept of a linear function, respectively. This linear function contacts a point on the curve corresponding to the open-circuit voltage-state-of-charge relationship, and this point is close to the first-order estimated state-of-charge value SOC. est - The state-of-charge data value N of (t) ε [SOC est - (t)] corresponds.

9. The method according to claim 1, wherein, The steps for calculating the current Kalman gain value include: Extract the first-order estimated state of charge (SOC) value from the coefficient data. est - The first coefficient value C(t) corresponding to (t); and By using the first coefficient value C(t) and the first-order estimated Kalman error covariance value P k - (t), the estimated G parameter value G est (t) and measurement noise σ v Based on L k (t)=C(t)×P k - (t) / [C(t) 2 ×P k - (t)+G est (t) 2 ×σ v To calculate the current Kalman gain value L k (t).

10. The method according to claim 1, wherein, By using the first-order estimated state of charge (SOC) value est - (t), the current Kalman gain value L k (t), the current voltage value V(t), and the estimated voltage value V est (t), based on SOC est (t) = SOC est - (t)+L k (t)×(V(t)-V est (t) is used to calculate the current SOC value. est (t).

11. The method according to claim 2, wherein, The steps for calculating the current Kalman error covariance value include: Extract the SOC value from the coefficient data and the first-order estimated SOC value. est - The first coefficient value C(t) corresponding to (t); and By using the first-order estimated Kalman error covariance value P k - (t), the current Kalman gain value L k (t), the first coefficient value C(t), the estimated G parameter value G est (t) and measurement noise σ v Based on P k (t)=P k - (t)-Lk(t) 2 ×[C(t) 2 ×P k - (t)+G est (t) 2 ×σ v To calculate the current Kalman error covariance value P k (t).

12. The method according to claim 1, wherein, The adaptive filter is a filter that uses the recursive least squares method.

13. The method of claim 1, further comprising generating estimated H-parameter values ​​from the current voltage value and the current current value using the adaptive filter, wherein, The estimated H-parameter value is a numerical value representing the effective potential determined by the local equilibrium potential distribution and resistance distribution in the battery.

14. The method according to claim 13, further comprising setting initial state vector values ​​and initial covariance matrix values ​​for the battery. in, The step of periodically generating the voltage and current values ​​of the battery includes: The previous voltage and current values ​​of the battery are generated; and The current voltage value and the current current value of the battery are generated after the sampling period Ts.

15. The method according to claim 14, wherein, The steps for generating the estimated G parameter values ​​and the estimated H parameter values ​​include: The estimated voltage value of the battery is calculated based on the current current value and the previous state vector value; The current gain matrix value and the current covariance matrix value are calculated based on the current current value and the previous covariance matrix value. The voltage error is calculated based on the current voltage value and the estimated voltage value; and The estimated G-parameter values ​​and the estimated H-parameter values ​​are generated by calculating the current state vector value based on the previous state vector value, the current gain matrix value, and the voltage error.

16. The method according to claim 15, wherein, By using the current current value I(t) and the previous G parameter value G est (t-1) and the previous H parameter value H est (t-1), based on V est (t)=G est (t-1)×I(t)+H est (t-1) is used to calculate the estimated voltage value V. est (t).

17. The method according to claim 15, wherein, By using the previous state vector value Θ est (t-1), the current gain matrix value L(t) and the voltage error e(t), according to Θ est (t)=Θ est The current state vector value Θ is calculated using (t-1)+L(t) ×e(t). est (t).

18. The method according to claim 15, wherein, The steps of generating the estimated G parameter value and the estimated H parameter value further include receiving a first forgetting factor λ1 associated with the G parameter and a second forgetting factor λ2 associated with the H parameter.

19. The method according to claim 18, wherein, The current gain matrix value is calculated using the following equation: ， The current covariance matrix value is calculated using the following equation: ， Where L(t) is the current gain matrix value, P(t) is the current covariance matrix value, P(t-1) is the previous covariance matrix value, I(t) is the current current value, λ1 is the first forgetting factor, and λ2 is the second forgetting factor.

20. The method according to claim 1, wherein, The steps of sensing the voltage and current of the battery for each predetermined sampling period Ts and periodically generating the voltage and current values ​​of the battery include: For each predetermined sampling period Ts, the voltage and current of the battery are sensed, and the sensed voltage value and sensed current value of the battery are generated; and The battery's voltage and current values ​​are periodically generated by inputting each of the sensed voltage and sensed current values ​​into a noise filter, and The method further includes generating estimated H-parameter values ​​from the voltage and current values ​​using the adaptive filter, wherein the estimated H-parameter values ​​are numerical values ​​of H-parameters representing the effective potential determined by the local equilibrium potential distribution and resistance distribution in the battery.

21. The method according to claim 20, wherein, The step of inputting the current current value from the periodically generated current values ​​and the current voltage value from the periodically generated voltage values ​​into the extended Kalman filter includes: Receive the current value as the current current value; and The sensed voltage value is received as the current voltage value.

Images