KR101792537B1 - Apparatus and method for battery status estimation, recording medium for performing the method - Google Patents

Abstract

The present invention relates to a method for estimating a battery state which estimates a state of charge (SOC) of a battery. The method comprises: calculating an open circuit voltage reflecting a diffusion phenomenon by using a diffusion index determined by a predetermined relaxation time; calculating an open circuit voltage analyzing a hysteresis phenomenon of the battery, and reflecting the hysteresis phenomenon from a modeled hysteresis loop; and estimating an SOC of the battery by applying the open circuit voltage reflecting the diffusion phenomenon and the open circuit voltage reflecting the hysteresis phenomenon to a dispersion point Kalman filter.

Description

BACKGROUND OF THE INVENTION 1. Field of the Invention [0001] The present invention relates to a battery state estimation device and method, and a recording medium for performing the same. BACKGROUND OF THE INVENTION [0002]

The present invention relates to an apparatus and method for estimating a battery state, a recording medium for performing the same, and more particularly, to an apparatus and method for estimating a battery state considering diffusion phenomenon and hysteresis phenomenon of a shielded flooded lead-acid battery, .

Energy storage systems are increasingly in demand from smart grids to plug-in hybrid electric vehicles and electric vehicles such as electric vehicles. Among energy storage systems, batteries are attracting attention in that they have a wider range of power, energy density and lower cost than other systems such as flywheels, ultracapacitors or hydrogen storage devices.

Such batteries are used in various types such as lithium ion (Li Ion), lead acid, and nickel cadmium. Of these, lead-acid batteries are widely used because of their low cost and long life.

Meanwhile, measuring the state of charge (SOC) of a battery in a battery management system (BMS) is one of the most basic requirements. The battery management system can use an appropriate equivalent model of the battery to protect the battery from overvoltage and overcurrent, and to measure the state of the battery, such as State of Charge (SOC) or State of Health (SOH).

At this time, the SOC-OCV relationship is a very important factor in modeling the battery equivalent circuit. Open circuit voltage (OCV) is calculated by SOC input and SOC-OCV relationship and can be combined with battery polarity. The SOC-OCV relationship is not a one-to-one mapping, but a hysteresis effect. That is, the OCV has a different value depending on the charge and discharge history even in the same SOC.

On the other hand, this hysteresis phenomenon is hardly taken into consideration in a lithium ion battery because the difference between SOC-OCV charge and discharge curves in most regions is very small. On the other hand, hysteresis phenomena should be considered very important for lead-acid batteries, especially shielded pledged lead-acid batteries. In other words, it is most important to model the hysteresis phenomenon in measuring the SOC of the lead-acid battery. The average charge and discharge curves are used to check the SOC-OCV relationship of the batteries that are not significantly affected by the hysteresis phenomenon. When checking the SOC-OCV relationship of the batteries, which are greatly affected by the hysteresis phenomenon in this way, . For example, there are a method of modeling a hysteresis curve by an experiment, a method using a first-order differential equation for expressing a voltage characteristic, a method using a parallelogram, and a method using a quadratic polynomial for modeling a hysteresis shape. All of these methods are based on the fact that the hysteresis loop always touches whenever the SOC-OCV discharge curve is applied to the SOC-OCV charge curve.

However, the conventional hysteresis phenomenon modeling methods are not suitable for use in a shielded flooded lead-acid battery, and a new hysteresis phenomenon modeling method suited to the characteristics of a shielded flooded lead-acid battery is needed.

One aspect of the present invention is a battery state estimating apparatus and method for estimating a battery state by calculating an open circuit voltage (OCV) reflecting a hysteresis phenomenon by modeling a hysteresis phenomenon of a shielded flooded lead-acid battery, Media.

According to an aspect of the present invention, there is provided a battery state estimation method for estimating a state of charge (SOC) of a battery, the method comprising: calculating an open circuit voltage reflecting diffusion phenomenon using a diffusion index determined according to a predetermined relaxation time; The open circuit voltage reflecting the hysteresis phenomenon and the open circuit voltage reflecting the hysteresis phenomenon are applied to the dispersion point Kalman filter to calculate the open circuit voltage reflecting the hysteresis phenomenon, Estimate the SOC.

On the other hand, calculating the open circuit voltage reflecting the diffusion phenomenon may further include calculating an open circuit voltage having an open circuit voltage having a mitigation time of 3 minutes and a mitigation time of 3 hours.

The calculation of the open circuit voltage reflecting the diffusion phenomenon may be performed between the open circuit voltage having the relaxation time of 3 minutes and the open circuit voltage having the relaxation time of 3 hours according to the diffusion index, Circuit voltage that reflects the diffusion phenomenon that is determined by the open circuit voltage.

Further, calculating the open circuit voltage reflecting the hysteresis phenomenon may include modeling a hysteresis loop of a parallelogram shape.

The calculation of the open circuit voltage reflecting the hysteresis phenomenon may be performed using the slope of the parallelogram of the hysteresis loop to calculate the open circuit voltage reflecting the hysteresis phenomenon.

The estimation of the SOC of the battery by applying the open-circuit voltage reflecting the diffusion phenomenon and the open-circuit voltage reflecting the hysteresis phenomenon to the distributed point Kalman filter may include estimating the SOC of the battery based on the open circuit voltage reflecting the diffusion phenomenon and the hysteresis phenomenon The SOC of the battery may be estimated by applying the open circuit voltage and the parameters of the battery to the dispersion point Kalman filter including prediction, observation, and measurement steps.

Further, the battery may be a shielded float lead-acid battery.

Further, it may be a computer-readable recording medium on which a computer program for performing the battery state estimation method is recorded.

According to another aspect of the present invention, there is provided a battery state estimating apparatus for estimating an SOC (State of Charge) of a battery, the battery state estimating apparatus estimating an SOC (State of Charge) of the battery using a diffusion index determined according to a predetermined relaxation time, A second OCV calculator for calculating an open circuit voltage reflecting the hysteresis phenomenon and an open circuit voltage reflecting the diffusion phenomenon and an open circuit voltage reflecting the hysteresis phenomenon to a dispersion point Kalman filter And an SOC estimation unit for estimating an SOC of the battery.

On the other hand, the first OCV calculator may further include calculating an open circuit voltage having an open circuit voltage having a relaxation time of 3 minutes and a relaxation time of 3 hours.

Further, the first OCV calculation unit calculates the first OCV based on the diffusion phenomenon determined between the open-circuit voltage having the relaxation time of 3 minutes and the open-circuit voltage having the relaxation time of 3 hours according to the diffusion index, The open circuit voltage can be calculated.

Further, the second OCV calculator may further include modeling a hysteresis loop of a parallelogram shape.

In addition, the second OCV calculator may calculate an open circuit voltage using the slope of the parallelogram of the hysteresis loop to display the hysteresis phenomenon.

Also, the SOC estimating unit may calculate the SOC of the battery based on the open circuit voltage reflecting the diffusion phenomenon, the open circuit voltage reflecting the hysteresis phenomenon, and the dispersion point including the prediction, observation, The SOC of the battery can be estimated using the Kalman filter.

Further, the battery may be a shielded float lead-acid battery.

According to an aspect of the present invention, a hysteresis loop matching a hysteresis characteristic of a shielded flooded lead-acid battery is modeled and a parameter calculated using the hysteresis loop is applied to a dispersion point Kalman filter UKF to estimate an SOC of the battery, Can be minimized.

1 is a control block diagram of a battery state estimation apparatus according to an embodiment of the present invention.
2 is an equivalent circuit diagram of a lead-acid battery.
3 is a graph showing an example of an SOC-OCV curve with 3 and 3 minute relaxation times.
FIGS. 4 to 6 are graphs for explaining the characteristics of the hysteresis phenomenon of the shielded flooded lead-acid battery.
7 is an example of a hysteresis loop modeled by the second OCV calculator.
FIG. 8 is a graph illustrating a result of estimating the SOC of a battery using a battery estimating apparatus according to an embodiment of the present invention and a result of estimating the SOC of the battery in a conventional manner.
9 is a flowchart illustrating a battery state estimation method according to an embodiment of the present invention.

The following detailed description of the invention refers to the accompanying drawings, which illustrate, by way of illustration, specific embodiments in which the invention may be practiced. These embodiments are described in sufficient detail to enable those skilled in the art to practice the invention. It should be understood that the various embodiments of the present invention are different, but need not be mutually exclusive. For example, certain features, structures, and characteristics described herein may be implemented in other embodiments without departing from the spirit and scope of the invention in connection with an embodiment. It is also to be understood that the position or arrangement of the individual components within each disclosed embodiment may be varied without departing from the spirit and scope of the invention. The following detailed description is, therefore, not to be taken in a limiting sense, and the scope of the present invention is to be limited only by the appended claims, along with the full scope of equivalents to which such claims are entitled, if properly explained. In the drawings, like reference numerals refer to the same or similar functions throughout the several views.

Hereinafter, preferred embodiments of the present invention will be described in more detail with reference to the drawings.

1 is a control block diagram of a battery state estimation apparatus according to an embodiment of the present invention.

1, a battery state estimation apparatus 100 according to an embodiment of the present invention includes a first OCV (Open Circuit Voltage) calculation unit 110, a second OCV calculation unit 120, and a State of Charge (SOC) ) Estimating unit 130 in accordance with the present invention.

The battery condition estimating apparatus 100 according to an embodiment of the present invention calculates an open circuit voltage (OCV) reflecting a diffusion phenomenon of a shielded flooded lead-acid battery, calculates an open circuit voltage (OCV) reflecting a hysteresis phenomenon, These open circuit voltages (OCVs) can be applied to a distributed point Kalman filter (UKF) to estimate the SOC of a shielded flooded lead-acid battery.

Particularly, the battery condition estimating apparatus 100 according to an embodiment of the present invention can estimate the SOC of the battery by modeling the hysteresis phenomenon by observing the hysteresis characteristic of the shielded flooded lead-acid battery, .

Hereinafter, each component of the battery condition estimation apparatus 100 shown in FIG. 1 will be described in detail.

First, the first OCV calculator 110 may calculate an open circuit voltage (OCV) reflecting the diffusion effect of the battery. This will be described with reference to FIG. 2 and FIG.

2 is an equivalent circuit diagram of a lead-acid battery.

2, an equivalent circuit of a battery is connected to an open circuit voltage OCV _f in series with a charge transfer resistor R _ct and a double layer capacitor C _dl , to which an internal resistor R _i is connected in parallel , And a voltage response due to the battery current. At this time, the open circuit voltage OCV _f may be the last stage open circuit voltage OCV including both diffusion and hysteresis.

On the other hand, the SOC-OCV curve showing the relationship between the SOC and the open circuit voltage (OCV) can be calculated with a relaxation time of 3 hours and 3 minutes. At this time, the relaxation time of 3 minutes can be selected because the OCV is almost fixed for 3 hours and the diffusion phenomenon of the charge transfer resistance (R _ct ) and the bilayer capacitor (C _dl ) is finished.

FIG. 3 is a graph showing an example of an SOC-OCV curve having a relaxation time of 3 hours and 3 minutes in this manner.

Referring to FIG. 3, it can be seen that the differences between the curves with the relaxation times of 3 minutes and 3 hours at different SOCs are different. At this time, the SOC-OCV curve having a relaxation time of 3 minutes and 3 hours can be expressed by a fifth-order polynomial.

On the other hand, the open circuit voltage (OCV) reflecting the diffusion phenomenon can be calculated using the following equation (1).

는 확산 지수를 의미하며 아래의 수학식 2를 이용하여 산출될 수 있다.OCV _d represents the open circuit voltage reflecting the diffusion phenomenon, OCV _3h and OCV _3m are the open circuit voltages having a relaxation time of 3 hours and 3 minutes respectively, s means SOC,

Means a diffusion index and can be calculated using the following equation (2).

는 확산 시간 상수, t는 완화 시간(단, 180s<t<10800s)을 의미한다. In Equation 2,

Is the diffusion time constant, and t is the relaxation time (note that 180s <t <10800s).

는 0과 1 사이에서 변할 수 있으며, 이에 따라 확산 현상을 반영한 개방 회로 전압(OCV_d)은 3분의 완화 시간을 갖는 개방 회로 전압(OCV_3m)과 3시간의 완화 시간을 갖는 개방 회로 전압(OCV_3h) 사이에서 변할 수 있다.According to equations (1) and (2)

(OCV _d ) reflecting the diffusion phenomenon can be varied between 0 and 1 so that the open circuit voltage (OCV _d ) having an open circuit voltage (OCV _{3 m} ) with a mitigation time of 3 min and an open circuit voltage OCV _3h. &_Lt; / RTI &gt;

Referring again to FIG. 1, the second OCV calculator 120 may model the hysteresis phenomenon of the shielded flooded lead-acid battery to calculate an open circuit voltage reflecting the hysteresis phenomenon.

The hysteresis phenomenon generally refers to a phenomenon in which the dependent variable is different from the forward path when the dependent variable changes back to the changed path. In terms of the electrochemical system, this means that the potential depends on the charge and discharge history instead of solely depending on the SOC. For example, the characteristics related to the hysteresis phenomenon in the nickel electrode are as follows. First, after the charging or discharging current is turned off, the hysteresis phenomenon in the excess time period of mass transport into the electrode is fixed, and the hysteresis phenomenon and the charge and discharge rate are independent, and finally, the hysteresis phenomenon occurs in the curve Lt; RTI ID = 0.0 &gt; sub-hysteresis loop. &Lt; / RTI &gt; The characteristics associated with this hysteresis phenomenon correspond to the domain theory, which is that the electrode material consists of two discrete regions of metastable state. The fludit battery also has such a hysteresis phenomenon characteristic, and the second OCV calculator 120 can model the hysteresis phenomenon of the shielded flooded lead-acid battery by reflecting the hysteresis developing characteristic.

FIGS. 4 to 6 are graphs for explaining the characteristics of the hysteresis phenomenon of the shielded flooded lead-acid battery.

Referring to FIG. 4, four hysteresis loops obtained using a 5% SOC pulse of 3 hours of relaxation time and SOC-OCV charge and discharge curves with a 3 hour relaxation time are shown and the OCV of all hysteresis loops Can not be seen to touch the SOC-OCV charge curve. In particular, it can be seen that a hysteresis loop with an SOC of 40-80% is approximately parallel to the SOC-OCV charge curve, which shows that the OCV of the hysteresis loop does not touch the SOC-OCV charge curve, even if the hysteresis loop is extended. The graph shown in FIG. 4 shows the characteristics of such a hysteresis loop of a shielded flooded lead-acid battery, whereas a battery such as a lead-acid AGM battery or a NiMH lithium ion battery has a hysteresis loop OCV is always in contact with the SOC-OCV charge curve.

Referring to FIG. 5, three hysteresis loops in the SOC range of 20% are shown, and it can be seen that the OCV increase rates in different SOC regions are different. This phenomenon is due to the fact that the OCV of the hysteresis loop of the shielded flooded lead battery does not touch the SOC-OCV charge curve, and thus it is desirable to model the hysteresis phenomenon of the shielded flooded lead battery to reflect this phenomenon.

Since the OCV of the hysteresis loop of the shielded flooded lead-acid battery does not touch the SOC-OCV charge curve, the second OCV calculator 120 can model the hysteresis loop without using the SOC-OCV charge curve. Each time the battery starts charging at a particular SOC value, the open circuit voltage (OCV) changes along the hysteresis loop.

Referring to FIG. 6, a hysteresis loop of three minutes relaxation time in a plurality of different SOC intervals is shown, and it can be confirmed that each hysteresis loop is a parallelogram shape. Accordingly, the second OCV calculator 120 can model the hysteresis phenomenon using the parallelogram.

7 is an example of a hysteresis loop modeled by the second OCV calculator 120. In FIG.

Referring to FIG. 7, it can be seen that the open circuit voltage (OCV) changes along the slope of k1 in the SOC interval of 0 to 5%, and the open circuit voltage (OCV) changes along the slope of k2 in the subsequent period . At this time, the slope in each section can be calculated using the following equation (3).

Equation 3 is a formula for obtaining the slope by dividing the amount of change in the y-axis by the amount of change in the x-axis in Fig. 7, that is, the amount of change in the open circuit voltage (OCV) Ah value).

Table 1 below is the slope of the hysteresis loop calculated when charging of the battery starts at different SOC values using Equation 3. &lt; EMI ID = 3.0 &gt;

Referring to FIG. 7, it can be seen that the slope at discharge also depends on the SOC value. Specifically, it can be confirmed that the slope at discharging and the slope at charging are matched in a specific SOC interval.

Thus, the relationship between the slope of the hysteresis loop and the SOC value at the charge / discharge start point is linear, and therefore, the open circuit voltage reflecting the hysteresis phenomenon can be calculated using Equation (4) below.

In Equation (4), OCV _hys is an open circuit voltage reflecting the hysteresis phenomenon, and k ₁ and k ₂ denote the slope of the hysteresis loop, respectively. C _n is the nominal capacity of the battery, and I is the battery current.

Referring again to FIG. 1, the SOC estimator 130 may estimate the SOC of the battery using the dispersion point Kalman filter UKF. At this time, the SOC estimating unit 130 can estimate the SOC of the battery by applying the battery open circuit voltage (OCV) parameter reflecting the diffusion phenomenon and the hysteresis phenomenon to the dispersion point Kalman filter UKF.

The distributed point Kalman filter (UKF) uses an unscented transformation instead of a Jacobian matrix for nonlinear systems, resulting in a more stable and better performance. Such a distributed point Kalman filter (UKF) can be largely composed of prediction, observation, and measurement steps. The equation of the discrete time step for the nonlinear system of the dispersion point Kalman filter (UKF) is shown in Equation (5) below.

In Equation (5), x _k denotes a state parameter, f denotes a non-linear system function, u _k denotes an input, w _k denotes a process noise vector, Y _k denotes an observed measurement signal, h denotes a non-linear system function, and v _k denotes a measurement noise vector.

The algorithm of the dispersion point Kalman filter (UKF) will be described in detail as follows. First, the data initialization step is expressed by Equation (6) below.

는 x₀의 기대값(expected value)을 나타내고, x₀는 state 'x'의 초기값을 나타내며, P₀는 초기 단계의 공분산 매트릭스(covariance matrix)를 나타낸다.In Equation (6)

Denotes the expected value of _{x 0 (expected value), x} 0 represents the initial value of the state 'x', P ₀ represents a covariance matrix (covariance matrix) of the initial stage.

The next step, Generate Sigma Points, is shown in Equation (7) below.

은 모든 시그마 포인트들을 포함하는 매트릭스를 나타내고,

은 'k-1'에서의 상태 값을 나타낸다. 그리고, 수학식 7에서

은 스케일링 지수(scaling factor)이고,

는 시그마 포인트의 분산(spread of sigma points)을 결정하게 되며 그 값은 10^-4와 1 사이에서 변할 수 있다. K는 일반적으로 '0'으로 설정될 수 있으며, L은 상태 벡터(state vector)의 숫자를 나타낸다.In Equation 7 chol is cholrae ski decomposition (decomposition cholesky) of the matrix, P _{k | k-1} denotes a value at a 'k-1' step of the covariance matrix,

Represents a matrix containing all sigma points,

Represents the state value at 'k-1'. In Equation 7,

Is a scaling factor,

Will determine the spread of sigma points, which can vary between 10 ^-4 and 1. K is generally set to '0', and L represents the number of state vectors.

In the next step, a prediction transformation step, the prediction state can be calculated by propagating the sigma point to the input function.

은 입력 함수 f에 시그마 포인트들을 전파시킨 뒤의 행렬을 나타내고, i는 시그마 포인트 넘버를 나타내며,

는 상태의 예측값(predicted value of the state)을 나타내고, -는 예측 업데이트(prediction update)를 나타내고,

평균 웨이트 매트릭스(mean weight matrix)를 나타낸다.In Equation (8)

Represents a matrix after propagating sigma points to an input function f, i represents a sigma point number,

Indicates a predicted value of the state, - indicates a prediction update,

Represents a mean weight matrix.

The propagated covariance can be calculated by the following equation (9).

는 공분산 매트릭스(covariance matrix)를 나타내고, 여기서 -는 예측 업데이트(prediction update)를 나타내며,

는 공분산 웨이트 매트릭스(covariance weight matrix), Qw 는 단계 노이즈(process noise)의 공분산 값을 나타낸다. In Equation (9)

Represents a covariance matrix, where - represents a prediction update,

Is the covariance weight matrix, and Qw is the covariance value of the process noise.

The next step, the observation transformation step, can propagate the sigma point to the output function to calculate the predicted measurement, as shown in Equation 10 below.

는 출력 함수 h 로 전파된 뒤의 시그마 포인트들을 포함하는 매트릭스를 나타내고,

는 출력 예측 값(predicted value of the output)을 나타낸다.In Equation (10)

Represents a matrix containing the sigma points that have propagated to the output function h,

Represents a predicted value of the output.

The estimated covariance can be calculated by Equation (11) below.

는 time step 'k'에서의 출력 공분산 매트릭스(output covariance matrix)를 나타내고, R_v는 측정 노이즈(measurement noise)의 공분산값을 나타낸다.In Equation (11)

Represents the output covariance matrix (covariance matrix output) at time step 'k', R _v represents the covariance value of the measured noise (noise measurement).

The cross covariance can be calculated by Equation (12) below.

는 time step 'k'에서의 교차 공분산 매트릭스(cross covariance matrix)를 나타낸다.In Equation 12,

Represents the cross covariance matrix at time step 'k'.

Finally, there is a measurement update step, in which the Kalman gain can be calculated as: &lt; EMI ID = 13.0 &gt;

및

는 각각 time step 'k'에서의 교차 공분산 매트릭스 및 출력 공분산 매트릭스를 나타낸다.In Equation 13 G _k denotes the Kalman gain (Kalman gain),

And

Denote the crossover covariance matrix and the output covariance matrix at time step 'k', respectively.

Then, the state and the covariance value can be updated as shown in Equation (14) below.

및 Pk 는 각각 time step 'k'에서의 상태 업데이트 값 및 공분산 매트릭스를 나타내고,

는 상태의 예측값(predicted value of the state),

는 출력 예측 값(predicted value of the output)을 나타내고, yk는 출력 측정 값(measured value of the output),

는 공분산 매트릭스(covariance matrix)를 나타낸다.In Equation (14)

And Pk denote the state update value and the covariance matrix at time step 'k', respectively,

A predicted value of the state,

Yk is a measured value of the output, yk is a measured value of the output,

Represents a covariance matrix.

The SOC estimating unit 130 can estimate the SOC of the battery using the dispersion point Kalman filter UKF having such a step. The state equation for this can be expressed by Equation (15) below.

In Equation 15, Δt represents a step time of the estimation algorithm, and T represents a transpose of matrix. V _Cdl is the voltage across the charge transfer resistor (R _ct ) and the bilayer capacitor (C _dl ) connected in parallel to the battery equivalent circuit, R _ct And C _dl are charge transfer resistances and double layer capacitors, respectively, of the battery equivalent circuit, C _n is the nominal capacity of the battery, R _i is the internal resistance of the battery equivalent circuit, I _k and V _k are the current Where w _k and v _k denote the process noise vector, and v _k denotes the measurement noise vector, respectively.

Also, OCV _d can be calculated from Equation 1 with an open circuit voltage reflecting diffusion phenomenon, and OCV _h can be calculated from Equation 4 with an open circuit voltage reflecting a hysteresis phenomenon.

Here, the battery parameters can be updated through the ARX method, which is described in &quot; Ngoc-Tham Tran, Kim-Hung Nguyen, Van-Long Pham and Woo-Jin Choi. 'SOC / SOH Estimation method for AGM VRLA battery by combining ARX model for online parameters estimation and DEKF considering hysteresis and diffusion effects', 9 th International Conference on Power Electronics and ECCA (ICPE-ECCA Asia), IEEE; Quot; June 2015. &quot;

Hereinafter, an advantageous effect of estimating SOC of a battery using the battery state estimation apparatus 100 according to an embodiment of the present invention will be described with reference to FIG.

FIG. 8 is a graph for comparing the SOC estimation result of the battery using the battery state estimation apparatus 100 according to an embodiment of the present invention and the SOC estimation result of the battery in a conventional manner.

 First, to obtain the graph shown in FIG. 7, experiments were conducted using a shielded pludded lead-acid battery having specifications of 90 Ah and 12 V. FIG. Referring to FIG. 7, the SOC estimation results (SOC-Variable Slopes) in the battery condition estimating apparatus 100 according to an embodiment of the present invention vary depending on the SOC of the battery every time the battery changes from the discharging state to the charging state .

Here, in the conventional system (SOC-Fixed Slopes), only one hysteresis model is used in all the SOC regions, and accordingly, as the hysteresis voltage increases, the SOC estimation error also increases and the error increases to 6% . That is, the conventional method is not suitable as the SOC estimation method of the shielded flooded lead-acid battery.

On the other hand, the SOC estimation error of the battery state estimation apparatus 100 according to an embodiment of the present invention is generally very accurate and is confirmed to be 3% at maximum.

As described above, the battery condition estimating apparatus 100 according to an embodiment of the present invention models a hysteresis loop matching a hysteresis characteristic of a shielded flooded lead-acid battery and applies the calculated parameters to the distributed point Kalman filter UKF So that the error can be minimized by estimating the SOC of the battery.

Hereinafter, a battery state estimation method according to an embodiment of the present invention will be described with reference to FIG.

9 is a flowchart illustrating a battery state estimation method according to an embodiment of the present invention.

The battery state estimating method according to an embodiment of the present invention can be performed in substantially the same configuration as the battery state estimating apparatus 100 shown in FIG. Therefore, the same components as those of the battery state estimation apparatus 100 shown in FIG. 1 are denoted by the same reference numerals, and a repeated description thereof will be omitted.

Referring to FIG. 9, the first OCV calculator 110 may calculate an open circuit voltage (OCV) reflecting the diffusion phenomenon of the battery (900). Here, the battery may be a shielded flooded lead-acid battery, the first OCV calculation unit 110 calculates an open circuit voltage having an open circuit voltage having a relaxation time of 3 minutes and a relaxation time of 3 hours, The open circuit voltage reflecting the diffusion phenomenon determined between the open circuit voltage having the relaxation time of 3 minutes and the open circuit voltage having the relaxation time of 3 hours is calculated according to the diffusion index which is a parameter reflecting the diffusion phenomenon can do.

The second OCV calculator 120 may calculate an open circuit voltage (OCV) reflecting the hysteresis of the battery (910). The shielded plodd lead-acid battery has confirmed that the OCV of the hysteresis loop does not touch the SOC-OCV charge curve even if the hysteresis loop is extended. Accordingly, the second OCV calculator 120 can detect the hysteresis loop without using the SOC- Loops can be modeled. The second OCV calculator 120 models a hysteresis loop of a parallelogram shape and calculates an open circuit voltage (OCV) reflecting the hysteresis phenomenon using the slope of the parallelogram as shown in [Equation 4] .

Finally, the SOC estimator 130 may estimate the SOC of the battery by applying the open circuit voltage reflecting the diffusion phenomenon and the hysteresis phenomenon to the dispersion point Kalman filter UKF. The SOC estimating unit 130 can estimate the SOC of the battery from Equation (15), and the battery parameters can be updated through the ARX method.

Such a battery state estimation method may be implemented in an application or may be implemented in the form of program instructions that can be executed through various computer components and recorded in a computer-readable recording medium. The computer-readable recording medium may include program commands, data files, data structures, and the like, alone or in combination.

The program instructions recorded on the computer-readable recording medium may be ones that are specially designed and configured for the present invention and are known and available to those skilled in the art of computer software.

Examples of computer-readable recording media include magnetic media such as hard disks, floppy disks and magnetic tape, optical recording media such as CD-ROMs and DVDs, magneto-optical media such as floptical disks, media, and hardware devices specifically configured to store and execute program instructions such as ROM, RAM, flash memory, and the like.

Examples of program instructions include machine language code such as those generated by a compiler, as well as high-level language code that can be executed by a computer using an interpreter or the like. The hardware device may be configured to operate as one or more software modules for performing the processing according to the present invention, and vice versa.

While the present invention has been particularly shown and described with reference to exemplary embodiments thereof, it will be understood by those skilled in the art that various changes and modifications may be made therein without departing from the spirit and scope of the invention as defined in the appended claims. It will be possible.

100: Battery condition estimating device

Claims (15)

A battery state estimation method for estimating an SOC (State of Charge) of a battery,
An open circuit voltage reflecting the diffusion phenomenon is calculated using a diffusion index determined according to a predetermined relaxation time,
The open circuit voltage reflecting the hysteresis phenomenon is calculated from the hysteresis loop modeled by analyzing the hysteresis phenomenon of the battery,
Applying an open circuit voltage reflecting the diffusion phenomenon, an open circuit voltage reflecting the hysteresis phenomenon and a parameter of the battery to a dispersion point Kalman filter composed of prediction, observation, and measurement steps, A method for estimating a battery state for estimating an SOC. The method according to claim 1,
The calculation of the open circuit voltage reflecting the diffusion phenomenon,
Further comprising calculating an open circuit voltage having an open circuit voltage having a relaxation time of 3 minutes and a relaxation time of 3 hours. 3. The method of claim 2,
The calculation of the open circuit voltage reflecting the diffusion phenomenon,
Calculating an open circuit voltage reflecting the diffusion phenomenon determined between an open circuit voltage having the relaxation time of 3 minutes and an open circuit voltage having the relaxation time of 3 hours in accordance with a diffusion index which is a parameter reflecting the diffusion phenomenon Of the battery state. The method according to claim 1,
The calculation of the open circuit voltage reflecting the hysteresis phenomenon,
And modeling a hysteresis loop of a parallelogram shape. 5. The method of claim 4,
The calculation of the open circuit voltage reflecting the hysteresis phenomenon,
And calculating an open circuit voltage reflecting the hysteresis phenomenon by using the slope of the parallelogram of the hysteresis loop. delete The method according to claim 1,
The battery includes:
A method for estimating battery condition as a shielded pledged lead-acid battery. A computer-readable recording medium on which a computer program is recorded, for performing the battery state estimation method according to any one of claims 1 to 5 and 7. A battery state estimation apparatus for estimating a state of charge (SOC) of a battery,
A first OCV (Open Circuit Voltage) calculating unit for calculating an open circuit voltage reflecting diffusion phenomenon by using a diffusion index determined according to a predetermined relaxation time;
A second OCV calculation unit for calculating an open circuit voltage reflecting a hysteresis phenomenon from a hysteresis loop modeled by analyzing a hysteresis phenomenon of the battery; And
And an SOC estimator for estimating an SOC of the battery by applying an open circuit voltage reflecting the diffusion phenomenon and an open circuit voltage reflecting the hysteresis phenomenon to a distributed point Kalman filter,
The battery includes:
A battery condition estimating device, which is a shielded pledged lead-acid battery. 10. The method of claim 9,
Wherein the first OCV calculation unit comprises:
Further comprising calculating an open circuit voltage having an open circuit voltage having a relaxation time of 3 minutes and a relaxation time of 3 hours. [Claim 11 is abandoned upon payment of the registration fee.] 11. The method of claim 10,
Wherein the first OCV calculation unit comprises:
A battery for calculating an open circuit voltage reflecting the diffusion phenomenon, which is determined between an open circuit voltage having the relaxation time of 3 minutes and an open circuit voltage having the relaxation time of 3 hours according to a diffusion index, State estimator. 10. The method of claim 9,
Wherein the second OCV calculation unit calculates,
Further comprising modeling a hysteresis loop of a parallelogram shape. [13] has been abandoned due to the registration fee. 13. The method of claim 12,
Wherein the second OCV calculation unit calculates,
And calculates an open-circuit voltage that has caused the hysteresis phenomenon by using the slope of the parallelogram of the hysteresis loop. 10. The method of claim 9,
The SOC estimator may include:
Wherein the battery voltage is measured by using the distributed point Kalman filter consisting of an open circuit voltage reflecting the diffusion phenomenon, an open circuit voltage reflecting the hysteresis phenomenon, and a parameter of the battery to predict, observe, and measure, And estimates the SOC of the battery.

delete

Images