KR100901252B1 - Method and Apparatus for estimation of State Of Charge using sliding mode observer - Google Patents

Abstract

The present invention relates to a method and apparatus for predicting state of charge (SOC) of a secondary battery using a sliding mode observer. Specifically, a secondary battery state prediction program for calculating a predicted value of the secondary battery output voltage (V _t ), the voltage across the capacitor (V _p ) and the SOC (Z) by a sliding mode designed observer, the parameters of the observer and The positive feedback gain constant and the open circuit voltage specific SOC table are read from the storage medium and loaded into the memory. Subsequently, the input current I and the output voltage V _t of the secondary battery are input from the current / voltage detection means at the sampling time t. Then the calculation algorithm of the observer is established by the loaded parameters and the positive feedback gain constant, and the SOC prediction value of the secondary battery is determined using the SOC table for each input current (I) and output voltage (V _t ) and open circuit voltage. Calculate and print According to the present invention, the SOC can be accurately predicted through simple circuit modeling even if the constant value of the secondary battery changes or there is an error in measuring the input current and / or output voltage of the secondary battery.

Secondary Battery, Sliding Mode Observer, SOC Prediction

Description

Method and apparatus for prediction of secondary battery SOC using sliding mode observer {Method and Apparatus for estimation of State Of Charge using sliding mode observer}

The following drawings attached to this specification are illustrative of the preferred embodiments of the present invention, and together with the detailed description of the invention to serve to further understand the technical spirit of the present invention, the present invention is a matter described in such drawings It should not be construed as limited to.

1 is a modeling circuit diagram of a secondary battery according to a preferred embodiment of the present invention.

2 is a block diagram of a hardware module constituting a sliding mode observer according to a preferred embodiment of the present invention.

3 is an external photograph of a secondary battery used in a constant current discharge experiment according to the present invention.

4 is a graph in which the open circuit voltage is plotted over the entire period of the state of charge (SOC) (0 ~ 1) using the results of the constant current discharge experiment according to the present invention.

FIG. 5 is a graph showing the current dependence of the polarization resistance R _p in the modeling circuit shown in FIG. 1.

6 shows the discharge current, the actual output voltage and the predicted output voltage of the secondary battery, and these output voltages for time t in the constant current discharge experiment of the on-line state using the sliding mode observer according to the preferred embodiment of the present invention. A graph plotting the error of the liver.

FIG. 7 is a graph in which the output voltage of the secondary battery and the actual value and the predicted value of the SOC are plotted together for a time t in the constant current discharge experiment using the sliding mode observer according to the preferred embodiment of the present invention.

FIG. 8 shows the input current, the actual output voltage and the predicted output voltage of these secondary batteries for time t in a 16-cycle Urban Dynamometer Driving Schedule (UDDS) experiment using a sliding mode observer according to a preferred embodiment of the present invention. Plot of error and SOC actual values.

FIG. 9 is an enlarged graph illustrating some sections of the graph illustrated in FIG. 8.

FIG. 10 is a graph plotting the prediction and actual values of SOC and an error between these values with respect to time t in a 16-cycle UDDS experiment using a sliding mode observer according to an exemplary embodiment of the present invention.

FIG. 11 is an enlarged graph illustrating some sections of the graph illustrated in FIG. 10.

The present invention relates to a method for predicting SOC of a secondary battery, and more particularly, to a method for accurately predicting SOC of a secondary battery using a sliding mode observer and an SOC prediction apparatus capable of implementing the method.

In recent years, hybrid electric vehicles (HEVs) with high energy efficiency and low combustion emissions have attracted attention as an alternative to new transportation. The major parts of hybrid electric vehicles are largely divided into internal combustion engines, electric motors and secondary batteries. The electric motor functions as a booster that receives energy from the secondary battery and starts the engine, and also functions as a generator when regenerating braking energy or supplying surplus power to the secondary battery for charging the secondary battery.

The control of the charge / discharge current of the secondary battery is based on the SOC which is the state of charge of the battery. The SOC information of the secondary battery is transmitted to a battery management system (BMS) through a control area network (CAN) communication line for power regulation of the electric motor.

SOC is nonlinear because it is determined by the chemical state inside the cell. Thus, SOC cannot be directly and accurately measured by an electrical signal, but indirectly predicted by physical measurements such as voltage across the battery, input current, and the like. In particular, large-capacity secondary batteries used for hybrid electric electric vehicles or electric vehicles having high C-rate are known to be difficult to accurately predict SOC because of their large nonlinearity.

Coulomb counting, impedance measurement, and Kalman filter are the most common methods for predicting SOC of a secondary battery.

The coulomb counting measures the actual amount of charge that can be released from a cell in '1 Ampere-hours'. For example, '1 Ampere-hours' indicates that the state of the battery is a state of charge capable of emitting 1 A of current for 1 hour. If an accurate current sensor is used, the Coulomb counting method must be a very accurate and inexpensive way of predicting SOC. However, since the SOC predictor employing the Coulomb counting method is an open loop predictor, an error generated in the current sensor accumulates by the SOC predictor, and thus, an error of the SOC increases with time.

Since the impedance measurement method is greatly influenced by the method and temperature for predicting the SOC of the battery from the measured impedance of the battery, there is a problem that the accuracy of the predicted SOC is lowered according to the temperature condition at which the impedance is measured.

Kalman filtering is one of the well-known techniques used for predicting the state of dynamic systems in the field of target tracking, navigation and so on. The Kalman filter method provides a recursive solution for optimal linear filtering for both observation and prediction of dynamic system states, and even if the system state is affected by global noise contained within the system's bandwidth. The advantage is that it can be optimally predicted.

US Pat. No. 6,534,954 discloses a method of predicting the charge of a cell using the Kalman filter method. However, the Kalman filter method has a disadvantage in that it is difficult to select the feedback gain. If the feedback gain is not properly selected, the predicted state of the system will diverge. In addition, the Kalman filter method has limitations in practical applications because system modeling must be done correctly and external noise must have a Gaussian distribution. If the above constraints are not met, the Kalman filter's performance will degrade and the system's condition will not be optimally predicted.

 Another method of predicting SOC is a method based on artificial neural network or fuzzy logic theory, as published in Korean Patent Laid-Open Publication No. 2005-61386. However, since these methods require highly complex mathematical operations, there is a problem of causing excessive load on a battery management system (hereinafter, referred to as 'BMS').

The present invention was devised to solve the above-mentioned problems of the prior art, and includes a sliding mode observer having robustness against disturbance, constant value change of a battery, and parameter change of a modeled secondary battery system. The purpose of the present invention is to provide a method and apparatus for accurately predicting SOC of a secondary battery.

Another object of the present invention is to provide a method and apparatus for accurately predicting SOC of a secondary battery even by system modeling of a simple secondary battery having uncertainty.

Another object of the present invention is to provide a method and apparatus for dynamically and accurately predicting SOC of a secondary battery using minimal parameters under various environments such as various temperatures and C-rates. .

,

및

를 계산하는 2차 전지 상태 예측 프로그램;을 수록한 저장매체를 이용하여 마이크로프로세서가 2차 전지의 SOC를 예측하는 방법이다. SOC prediction method of a secondary battery using a sliding mode observer according to the present invention for achieving the above technical problem, the capacitor component (C) by the open circuit voltage (V _oc (Z)) of the secondary battery, the polarization effect of the battery _p ) and the parameters of the respective state equations for the secondary cell output voltage (V _t ), capacitor across voltage (V _p ) and SOC (Z) derived by RC circuit modeling taking into account the resistance components inside the cell; A positive feedback gain constant of the observer designed in sliding mode based on each state equation; Open circuit voltage-specific SOC tables; And estimates of V _t , V _p, and Z based on the observer

,

And

The secondary battery state prediction program for calculating the; is a method of predicting the SOC of the secondary battery using a storage medium containing the microprocessor.

를 계산하여 출력한다.Specifically, first, the SOC table for each program, parameter, positive feedback gain constant, and open circuit voltage is loaded from the storage medium into a memory. Then, at the sampling time t, the input current I and the output voltage V _t of the secondary battery are input from the current / voltage detection means. Then, the equation of the observer is established by the loaded parameters and the positive feedback gain constant, and the SOC prediction value of the secondary battery using the SOC table for each input current (I) and output voltage (V _t ) and open circuit voltage.

Calculate and output

Preferably, the resistance component comprises a ripple resistance component R _b , a diffusion resistance component R _p and an ohmic resistance component R _t .

According to the present invention, each state equation for the secondary battery output voltage (V _t ), the voltage across the capacitor (V _p ) and SOC (Z) is represented by the following equation,

,

,

,

,

,

,

로서, 다음과 같이 2차 전지의 RC 상수값에 의해 계산된다. The parameter contained in the storage medium is a coefficient of the state equation.

,

,

,

,

,

,

It is calculated by the RC constant value of the secondary battery as follows.

And, the sliding mode observer for the secondary battery output voltage (V _t ), the voltage across the capacitor (V _p ) and SOC (Z) is represented by the following equation,

,

및

이다.The positive feedback gain constant stored in the storage medium is defined by the sliding mode observer equation.

,

And

to be.

,

및

의 계산을 위한 수학적 알고리즘을 확립하고,Preferably, the step of calculating and outputting the predicted value of the SOC, the output voltage (V _t ) of the secondary battery, the voltage across the capacitor (V _t ) by first substituting the parameters and positive feedback gain constant stored in the storage medium to the following equation: _p ) and SOC (Z) estimates

,

And

Establish a mathematical algorithm for the calculation of

,

및

를 계산하여 출력하는 단계이다.The input voltage (I) and output voltage (V _t ) input at the sampling time t, and the open circuit voltage and SOC at the time t contained in the SOC table for each open circuit voltage are substituted into the discrete equations to output the secondary battery output voltage. , Estimate of voltage across capacitor and SOC

,

And

It calculates and outputs it.

Preferably, the calculated SOC prediction is input to a battery management system (BMS).

와 실제치

사이의 오차는 3% 이내이다.Preferably, the SOC prediction

And actual value

The error between is within 3%.

In some cases, the storage medium may further include a capacitor component value list and a resistance component value list according to the input current magnitude of the secondary battery. In this case, calculating and outputting the predicted value of the SOC, the capacitor component value and the resistance component value according to the magnitude of the detected input current at the time t when the input current of the secondary battery is detected from the list to read the parameter The calculation is performed in real time, and the parameters used in the establishment of a mathematical algorithm for calculating the SOC prediction are the parameters calculated in real time.

,

및

를 계산하는 수학적 알고리즘을 포함하는 2차 전지 상태 예측 프로그램;을 수록하고 있는 저장매체; 샘플링 타임 t에서 이차 전지의 입력 전류(I)와 출력 전압(V_t)을 측정하여 출력하는 전류/전압 검지 수단; 및 상기 저장매체로부터 상기 2차 전지 상태 예측 프로그램, 상기 파라미터, 상기 정 궤환 이득 상수 및 상기 개방 회로 전압별 SOC 테이블을 메모리에 로드하는 한편, 상기 전류/전압 검지 수단으로부터 출력된 이차 전지의 입력 전류 및 출력 전압을 입력받아 상기 로드된 프로그램의 수학적 알고리즘에 따라 상기 파라미터, 정 궤환 이득 상수 및 개방 회로 전압별 SOC 테이블을 이용하여 2차 전지의 SOC 예측치

를 계산하여 출력하는 마이크로프로세서;를 포함한다.The SOC prediction apparatus of a secondary battery using the sliding mode observer according to the present invention for achieving the above another technical problem, the capacitor component by the open circuit voltage (V _oc (Z)), the polarization effect of the battery ( C _p ) and the parameters of the respective state equations for the secondary cell output voltage (V _t ), capacitor across voltage (V _p ) and SOC (Z) derived by RC circuit modeling taking into account the resistance components inside the cell; A positive feedback gain constant of the observer designed in sliding mode based on each state equation; Open circuit voltage-specific SOC tables; And estimates of V _t , V _p, and Z based on the observer

,

And

A secondary battery state prediction program comprising a mathematical algorithm for calculating a; storage medium containing; Current / voltage detecting means for outputting the sampling time t by measuring the input current (I) and output voltage (V _t) of the secondary battery; And loading the secondary battery state prediction program, the parameter, the positive feedback gain constant, and the open circuit voltage-specific SOC table from the storage medium into a memory, and inputting the secondary battery output from the current / voltage detection means. And an SOC prediction value of a secondary battery using the SOC table for each parameter, positive feedback gain constant, and open circuit voltage according to a mathematical algorithm of the loaded program.

It includes; a microprocessor for calculating and outputting.

Hereinafter, exemplary embodiments of the present invention will be described in detail with reference to the accompanying drawings. Prior to this, terms or words used in the specification and claims should not be construed as having a conventional or dictionary meaning, and the inventors should properly explain the concept of terms in order to best explain their own invention. Based on the principle that it can be defined, it should be interpreted as a concept and concept corresponding to the technical idea of the present invention. Therefore, the embodiments described in the specification and the drawings shown in the drawings are only the most preferred embodiment of the present invention and do not represent all of the technical idea of the present invention, various modifications that can be replaced at the time of the present application It should be understood that there may be equivalents and variations.

Sliding mode  Theoretical Background of Observers>

Prior to describing the preferred embodiment of the present invention in detail, the theoretical background of the sliding mode observer employed in the present invention will be described.

First, consider a continuous-time single input system and a scalar measurement model mathematically modeled by Equations 1 and 2 below.

는 모델링 된 시스템의 시간 t에서의 상태(sate variable)이고,

는 시간 t에서의 스칼라 궤환 제어에 따른 시스템 제어 입력이고,

는 시간 t에서의 경계가 있는(bounded) 오차 및 왜란(disturbance)을 정량화한 인 자이고,

은 시간 t에서의 시스템 출력이다. 그리고,

이고,

이고,

이다.From here,

Is the state variable at time t of the modeled system,

Is the system control input according to scalar feedback control at time t,

Is a factor that quantifies bounded errors and disturbances at time t,

Is the system output at time t. And,

ego,

ego,

to be.

에 대한 슬라이딩 모드 관측기는 다음 수학식 3과 같이 정의된다. 수학식 3에서,

은 각각

의 예측치(estimate)이다.In the system modeled by Equations 1 and 2 above, states

The sliding mode observer for is defined as in Equation 3 below. In Equation 3,

Are each

Is an estimate of.

이고,

이고,

는 시그넘(signum) 함수이고,

은 스위칭 궤환 이득 상수이다. here,

ego,

ego,

Is the signum function,

Is the switching feedback gain constant.

는 하기 수학식 4와 같은 동적 특성을 갖는다. Error in Sliding Mode Observer

Has dynamic characteristics as shown in Equation 4 below.

이고, here,

ego,

는

이다.Switching function

Is

to be.

이라는 조건이 성립되면 국소적인 슬라이딩 레자임(regime)은 시스템 상태 공간 내의

이라는 평면 상에 있게 된다. On the other hand, according to the sliding mode theory,

If a condition is established, the local sliding regime is defined in the system state space.

Is on the plane of.

는 하기 수학식 5로 나타낼 수 있다. 수학식 5에서 T는 전치(transpose) 행렬을 의미한다. Meanwhile above

Can be represented by Equation 5 below. In Equation 5, T means a transpose matrix.

이 0보다 크면, 국소적인 슬라이딩 레자임은 조건 In Equation 5,

If this is greater than zero, the local sliding regime is a condition

하에서

라는 평면 위에 존재하게 된다. 이때 집합

는 슬라이딩 패치(patch)라고 불린다.

Under

Will exist on the plane. At this time

Is called a sliding patch.

는 'Filippov'의 근(solution)에 의해 결정된다. On the other hand, in the ideal sliding mode dynamics,

Is determined by the solution of 'Filippov'.

에 대하여,That is, the condition

about,

이고,

이다.

ego,

to be.

는 스위칭 함수를

의 스팬(span)을 따라

의 널 공간(null space)으로 투영했을 때의 행렬이다.here,

The switching function

Along the span of

The matrix when projecting into null space.

The theoretical background of the sliding mode observer described above will be referred to in the embodiments of the present invention described below. However, it will be apparent to those skilled in the art that various well-known sliding mode observer theories may be applied in addition to the theories presented herein.

Sliding mode  Of secondary battery for the design of observer modelling >

Next, a modeling circuit of the secondary battery employed in the present invention will be described in order to predict the SOC of the secondary battery according to a preferred embodiment of the present invention.

In general, the output voltage of the secondary battery has a characteristic of changing nonlinearly with respect to the SOC of the battery. Such nonlinear characteristics can be easily confirmed by measuring the open-circuit voltage of the secondary battery while periodically charging the secondary battery and then periodically discharging a certain amount of current.

The present inventors dynamically modeled the secondary battery by the RC circuit model as shown in FIG. 1 in view of the nonlinear characteristics of the secondary battery as described above.

(참고로, Z는 SOC), 전지 내의 분극(polarization) 효과를 모델링하기 위한 캐패시터 성분

, 파급 저항(propagation resistance)을 모델링하기 위한 파급 저항 성분

, 전류(I)의 함수인 확산 저항(diffusion resistance) 성분

, 오믹 저항 성분

및 2차 전지의 출력 전압

를 포함한 다. 도 1에서, 캐패시터

의 양단 전압은

로 표시하였다.Referring to FIG. 1, a modeling circuit for a secondary battery includes an open circuit voltage exhibiting nonlinear characteristics.

(For reference, Z is SOC), a capacitor component for modeling the polarization effect in the cell

Ripple component for modeling propagation resistance

, Diffusion resistance component as a function of current (I)

Ohmic resistance components

And output voltage of secondary battery

It includes. In Figure 1, a capacitor

The voltage at both ends of

Marked as.

는 하기 수학식 6 및 7로 표시할 수 있다. 하기 수학식 6 및 7에서, I는 순시 전류(instantaneous current)이다. 순시 전류는 충전일 경우 '양'의 값을, 방전일 경우 '음'의 값을 갖는다.Output voltage of secondary battery

Can be represented by the following equations (6) and (7). In Equations 6 and 7, I is instantaneous current. Instantaneous current has a positive value for charging and a negative value for discharging.

On the other hand, the time derivative of SOC Z of the secondary battery is expressed by Equation 8 below.

는 개방 회로 전압의 전류이고

는 2차 전지의 공칭 정전용량(nominal capacitance)이다.here,

Is the current of the open circuit voltage

Is the nominal capacitance of the secondary battery.

Since the left sides of Equations 6 and 7 are identical to each other, the following Equation 9 can be derived through a simple algebra operation.

If Kirchoff's law of Equation 10 is applied to Equation 9, Equation 11 can be obtained.

Substituting Equation 11 into Equation 8 provides Equation 12 below.

By substituting Equation 12 into Equation 10 in a similar manner, Equation 13 is obtained.

는 수학식 11의

를 수학식 7에 대입하면 하기 수학식 14와 같이 얻을 수 있다. Output voltage of secondary battery

Of equation (11)

By substituting into Equation 7, it can be obtained as Equation 14 below.

이라는 조건을 적용하여 정리하면, 하기 수학식 15가 얻어진다.Time differential the output voltage according to Equation 14,

If the following condition is applied and summarized, the following formula (15) is obtained.

에 대하여 풀고 그 결과를 수학식 14에 대입하면, 하기 수학식 16과 같이

,

및

에 대한 완전한 상태 방정식을 얻을 수 있다. 이로써, 본 발명에 따른 2차 전지의 회로 모델링이 완료된다.Finally, Equation 6

Solving for and substituting the result into Equation 14,

,

And

You can get the complete state equation for. This completes the circuit modeling of the secondary battery according to the present invention.

,

,

,

,

,

,

는 다음과 같다.here,

,

,

,

,

,

,

Is as follows.

,

,

,

,

,

,

의 계산을 위해 사용되는 2차 전지의 상수값

,

,

,

,

는 2차 전지를 만 충전시킨 후 주기적으로 정 전류를 방전시키면서 개방 회로 전압을 측정한 후 그 실험 결과와 매칭될 수 있도록 시행착오법에 의해 상기 상수값들을 결정한다.In the above state equation, the parameter

,

,

,

,

,

,

Constant value of secondary battery used for calculation of

,

,

,

,

After the battery is charged only, the open circuit voltage is measured while periodically discharging a constant current, and the constant values are determined by a trial and error method so as to match the experimental result.

Sliding mode  Observer Design>

The design of the sliding mode observer according to the preferred embodiment of the present invention is based on the state equation according to Equation 16 above. Specifically, the rank of the observability matrix of Equation 16 is full rank. Therefore, the internal state of the secondary battery modeled by the circuit shown in FIG. 1 can be predicted by the sliding mode observer.

For reference, the observability matrix is [C CA CAA].

이고, A는

이다. Where C is

And A is

to be.

에 대한 상태 방정식에 적용하면,

에 대한 슬라이딩 모드 관측기는 하기 수학식 17과 같이 나타낼 수 있다.In the present invention, the design of the sliding mode observer starts from the equations (3) and (5) described above. Specifically, Equation 3 is expressed in Equation 16

Applying to the state equation for,

The sliding mode observer for may be represented by Equation 17 below.

은

의 예측치이고,

은 정 궤환 이득 상수(positive feedback gain constant)이다. here,

silver

Is an estimate of,

Is the positive feedback gain constant.

를

로 정의하면, 하기 수학식 18과 같은 오차 방정식을 얻을 수 있다.Estimation and Actual Error of System Output

To

If defined as, the error equation shown in Equation 18 can be obtained.

는 다음과 같다.here,

Is as follows.

이 충분히 크면

에 의해

의 부호가 결정된다. 그리고

에 의해

와

의 부호는 항상 반대가 된다. 그 결과,

값에 상관없이

에 슬라이딩 모드 운동이 유발됨으로써 일정한 시간이 흐른 뒤에는

및

가 0으로 수렴한다. 여기서, 상기 정 궤 환 이득 상수

은 시행착오법(trial and error)에 의해 결정된다.Referring to Equation 18,

If this is big enough

By

The sign of is determined. And

By

Wow

Is always reversed. As a result,

Regardless of the value

After a certain period of time,

And

Converges to zero. Where the positive feedback gain constant

Is determined by trial and error.

가 그 등가 치인

로 치환된 것과 같은 양상을 보인다. 이때,

는 수학식 18에서

및

가 0이라고 가정하여 구하므로

와

사이에는 하기 수학식 19와 같은 관계가 성립한다.Meanwhile, according to the equivalent control method, the error system in the sliding mode

That is equally hit

It shows the same aspect as substituted with. At this time,

In Equation 18

And

Is assumed to be 0, so

Wow

The following relationship holds for the following equation (19).

에 대한 슬라이딩 모드 관측기의 설계가 완료된다.Output voltage of the secondary battery by derivation of the relationship according to Equation 19

The design of the sliding mode observer for is completed.

에 대한 슬라이딩 모드 관측기 설계 방법을 설명한다. Next SOC

A sliding mode observer design method is described.

에 대한 상태 방정식에 적용하면,

에 대한 슬라이딩 모드 관측기는 하기 수학식 20과 같이 나타낼 수 있다.Specifically, Equation 3 is replaced by Equation 16

Applying to the state equation for,

The sliding mode observer for may be represented by Equation 20 below.

은

의 예측치이고,

정 궤환 이득 상수이다. here,

silver

Is an estimate of,

Positive feedback gain constant.

및

의 오차

및

를 각각

및

로 정의하면, 하기 수학식 21과 같은 오차 방정식을 얻을 수 있다.SOC

And

Error

And

Each

And

If defined as, an error equation such as the following Equation 21 can be obtained.

는 다음과 같다.here,

Is as follows.

는 비선형적이긴 하지만 대체로 SOC

에 비례한다. 따라서

는 하기 수학식 22와 같이 구간 구간 별로

에 비례한다고 근사할 수 있다.Open circuit voltage

Is non-linear, but usually SOC

Proportional to therefore

Is for each section as shown in Equation 22 below.

It can be approximated to be proportional to.

Therefore, Equation 21 may be arranged as Equation 23 below.

의 오차 방정식(수학식 18)과 마찬가지로,

가 충분히 크면

및

의 부호는 항상 반대가 되므로

에 슬라이딩 모 드 운동이 유발됨으로써 일정한 시간이 흐른 뒤에는

및

모두가 0으로 수렴한다. 여기서, 상기 정 궤환 이득 상수

는

과 마찬가지로 시행착오법에 의해 결정된다.Referring to Equation 23,

As with the error equation (Equation 18),

Is large enough

And

Is always reversed.

After a certain period of time,

And

All converge to zero. Where the positive feedback gain constant

Is

Similarly, it is determined by the trial and error method.

에 대한 슬라이딩 모드 관측기의 설계시와 마찬가지로 등가 제어 방법론을 적용하면, 하기 수학식 24와 같은 관계식을 도출할 수 있다.Also

By applying the equivalent control methodology as in the design of the sliding mode observer for the relation, it is possible to derive the relational expression as shown in Equation (24).

에 대한 슬라이딩 모드 관측기의 설계가 완료된다. If the relation according to Equation 24 is derived, SOC of the secondary battery

The design of the sliding mode observer for is completed.

에 대한 슬라이딩 모드 관측기와 오차 방정식을 설계하면 하기 수학식 25 및 26과 같다.Finally, apply the above process substantially the same

When the sliding mode observer and the error equation is designed for Equations 25 and 26.

가 충분이 크면

및

의 경우와 마찬가지로, 일 정한 시간이 경과되면

및

는 0으로 수렴한다. 상기 정 궤환 이득 상수

는

와 마찬가지로 시행착오법에 의해 결정된다.In Equation 26,

Is big enough

And

As in the case of,

And

Converges to zero. The positive feedback gain constant

Is

Similarly, it is determined by the trial and error method.

Sliding mode  Configuration and Operation of Observer>

The sliding mode observer according to the present invention is implemented as a hardware module electrically coupled with the secondary battery as shown in FIG. 2.

Referring to the drawings, the sliding mode observer 100 includes a current sensor 110, a voltage sensor 120, a microprocessor 130, a memory 140, a storage medium 150, an A / D converter 160 and I. / O interface 170 is included.

The current sensor 110 measures the current I input to the secondary battery B and inputs it to the A / D converter 160 in the form of an analog signal. The voltage sensor 120 measures the output voltage V _t of the secondary battery B and inputs it to the A / D converter 160 in the form of an analog signal. The A / D converter 160 then converts the input analog voltage and current signals into digital signals and inputs them to the microprocessor 130.

,

,

,

,

,

,

와, 이들 파라미터 계산에 사용되는 RC 성분인

,

,

,

,

와, 상태 방정식의 정 궤환 이득 상수

,

및

와, 정 전압 방전 실험을 통하여 획득한 개방 회로 전압별 SOC 값을 수록하고 있는 테이블과, 상기 파라미터; 상기 정 궤환 이득 상수; 상기 개방 회로 전압별 SOC 테이블; 상기 전류 센서(110) 및 전압 센서(120)에 의해 수집되는 2차 전지(B)의 입력 전 류(I) 및 출력 전압(V_t)을 이용하여 주기적으로 SOC의 예측치

를 출력하는 SOC 예측 프로그램이 수록되어 있다. 상기 저장매체는 불활성 메모리인 것이 바람직하고, 예컨대 플래쉬 메모리, ROM, EEPROM 등이 채용될 수 있는데, 본 발명이 이에 한하는 것은 아니다.The storage medium 150 is a parameter used in the state equation of the secondary battery B.

,

,

,

,

,

,

And the RC component used to calculate these parameters

,

,

,

,

And the positive feedback gain constant of the state equation

,

And

And a table for storing SOC values for each open circuit voltage obtained through the constant voltage discharge experiment, and the parameters; The positive feedback gain constant; An SOC table for each of the open circuit voltages; The prediction value of the SOC periodically using the input current I and the output voltage V _t of the secondary battery B collected by the current sensor 110 and the voltage sensor 120.

The SOC prediction program that outputs is shown. Preferably, the storage medium is an inert memory. For example, a flash memory, a ROM, an EEPROM, or the like may be employed, but the present invention is not limited thereto.

The microprocessor 130 controls the overall operation of the sliding mode observer 100. The microprocessor 130 executes the SOC prediction program contained in the storage medium 150 in order to proceed with the initialization of the sliding mode observer 100 when power is applied, while the secondary battery state equation (see Equation 16) The parameter value and the positive feedback gain constant are loaded into the memory 140 and the initial current and voltage data input from the A / D converter 160 are stored in the memory 140.

Here, the initial currents and voltages refer to currents and voltage levels measured just before starting of the vehicle on which the sliding mode observer 100 is mounted (before ignition starts). Therefore, the initial current is 0, and the initial voltage has a predetermined value according to the charge amount of the secondary battery B. In this case, the initial voltage is regarded as an initial open circuit voltage of the secondary battery B for convenience.

,

및

를 계산하여 메모리(140)에 누적적으로 저장하는 한편, SOC의 예측치

를 외부로 출력한다. Meanwhile, the microprocessor 130 reads the SOC value corresponding to the initial voltage value by referring to the SOC table for each open circuit voltage stored in the memory 140. Then periodically

,

And

Is calculated and stored cumulatively in the memory 140, while the estimated value of the SOC

To the outside.

,

및

를 계산한다. Specifically, in the present invention, by applying the above-described equivalent control methodology to the equation 17, 20 and 25 of the sliding mode observation equation of the secondary battery (B) to convert the sliding mode observation equation to Euler discrete type as shown in Equation 27 Sampling time T _s

,

And

Calculate

,

및

의 구체적인 계산 방식은 다양하게 변형이 가능하며, 본 발명의 기술적 범위가

,

및

의 구체적인 계산 방식에 의해 한정되지 않음은 당연하다. In an embodiment of the invention, the sampling time T _s Is 1 second. On the other hand, the size of the sampling time,

,

And

The specific calculation method of the can be variously modified, the technical scope of the present invention

,

And

Of course, it is not limited by the specific calculation method of.

,

및

를 계산하기 위해 상기 수학식 27의 초기 조건을 다음과 같이 설정한다.The microprocessor 130 of the secondary battery (B)

,

And

In order to calculate the initial condition of Equation 27 is set as follows.

은 차량의 시동 전에 측정된 2차 전지(B)의 초기 전압 레벨로,

은 차량의 시동 전에 측정된

를 개방 회로 전압

로 간주하여 메모 리(140)에 저장된 개방 회로 전압별 SOC 테이블을 참조하여 얻은 2차 전지(B)의 초기 SOC 값으로,

은 차량의 시동 전에 흐르는 전류 I가 0이라는 점을 감안하여

과 동일한 값으로,

;

;

는 계산상의 편의를 위해

, 0,

으로 초기 상태를 설정한다. 슬라이딩 모드 관측기의 특성상 초기 조건

;

;

를 위와 같이 설정해도 관측기의 성능에 유의적인 영향을 미치지 않는다.In other words,

Is the initial voltage level of the secondary battery (B) measured before starting of the vehicle,

Is measured before the vehicle starts up.

Open circuit voltage

As the initial SOC value of the secondary battery B obtained by referring to the SOC table for each open circuit voltage stored in the memory 140,

Considering that the current I flowing before starting the vehicle is zero.

Is the same value as

;

;

For convenience of calculation

, 0,

Set the initial state with. Initial conditions due to the nature of the sliding mode observer

;

;

Setting as above does not significantly affect the performance of the observer.

,

,

을 계산하여 메모리(140)에 저장하고

는 I/O 인터페이스(170)를 통해 외부로 출력한다.When the initial state is set, the microprocessor 130 applies the above initial condition to Equation 27 before the next time sampling time (t = 2 seconds) is reached, and estimates at t = 2 seconds.

,

,

Calculate and store in memory 140

Outputs to the outside through the I / O interface 170.

,

,

와 샘플링 타임(t=2초)에서 측정된

와 I, 개방 회로 전압별 SOC 테이블에서

에 상응하는

에 의하여

,

,

을 계산하고 I/O 인터페이스(170)을 통해

를 외부로 출력한다. After that, the microprocessor 130 stores the memory 140 when the next time sampling time (t = 2 seconds) is reached.

,

,

And measured at sampling time (t = 2 seconds)

And I, in the SOC table by open circuit voltage

Equivalent

By

,

,

Is computed and through the I / O interface 170

To the outside.

,

및

에 대한 계산 과정은 샘플링 주기 T_s가 경과될 때마다 반복적으로 이루어진다. Of the secondary battery (B)

,

And

The calculation process for is repeated every time the sampling period T _s has elapsed.

는 전지 관리 시스템(BMS)에 입력된다. 그러면 전지 관리 시스템(BMS)은 SOC의 예측치

에 의해 2차 전지(B)의 내부 상태를 판정하고, 만약 충전이 필요한 경우(예컨대

값이 소정 레벨 이하로 떨어진 경우)는 전정류/정전압 모드에서 2차 전지(B)를 충전시킨다. 전지 관리 시스템(BMS)에 의한 2차 전지 충전 동작은 본 발명이 속한 기술분야에서 널리 알려진 기술이므로 여기에서의 상세한 설명은 생략하기로 한다.Preferably, the predicted value of the SOC output through the sliding mode observer 100

Is input to the battery management system (BMS). The battery management system (BMS) then estimates the SOC

Determine the internal state of the secondary battery B, and if charging is necessary (e.g.,

Value drops below a predetermined level), the secondary battery B is charged in the pre-constant / constant voltage mode. Since the secondary battery charging operation by the battery management system (BMS) is well known in the art to which the present invention pertains, a detailed description thereof will be omitted.

,

,

,

,

,

,

의 계산을 위해 사용되는 RC 성분

,

,

,

,

는 상수일 수도 있지만, 2차 전지(B)의 입력 전류 및/또는 SOC 에 의해 변화될 수도 있다. 이러한 경우, 상기 저장매체(150)는 입력 전류 및/또는 SOC의 변화에 따라 실험적으로 결정된 RC 성분 값들의 테이블을 추가로 수록하고 있을 수 있다. 그러면 마이크로프로세서(130)는 샘플링 시간 T_s가 경과될 때마다 저장매체(150)에 수록된 RC 성분들의 테이블을 참조하여 파라미터

,

,

,

,

,

,

의 계산을 실시간으로 수행할 수 있다. On the other hand,

,

,

,

,

,

,

RC component used for calculation of

,

,

,

,

May be a constant, but may be changed by the input current and / or SOC of the secondary battery B. In this case, the storage medium 150 may further include a table of RC component values experimentally determined according to the change of the input current and / or SOC. The microprocessor 130 may then refer to the table of RC components contained in the storage medium 150 every time the sampling time T _s elapses.

,

,

,

,

,

,

Can be calculated in real time.

Experimental Example

The present inventors adopted a secondary battery as shown in FIG. 3 as a test cell to test the performance of the sliding mode observer designed according to the present invention. The secondary battery employed is a lithium-polymer secondary battery with negative and positive electrodes of LiMn ₂ O ₄ and graphite, respectively, with a nominal capacity of 5.0 Ah, a nominal voltage of 3.8 V, and a dimension of '250 x 125 x 5 [ mm] 'and the weight of the battery was 120 [g].

,

,

,

,

,

,

를 결정하고 전지의 개방 회로 전압을 SOC의 전 구간(0 ~ 1)에 대해 설정하기 위한 목적으로 2차 전지를 4.2V까지 완충전시킨 후 정 전류 방전 실험을 수행하였다. 정 전류 방전 실험은 180초 동안 5A의 전류를 방전시켰다가 3600초 동안 방전을 휴지하는 과정을 하나의 싸이클로 하여 20회 반복하였다. 참고로, 180초 동안 5A의 전류를 방전시키는 것은 공칭 캐패시티 5.0Ah의 '1-C rate'에 해당한다. 그리고 상기 정 전류 방전 실험에서 5A의 전류가 방전되면 5%의 SOC가 감소된다. 본 발명자는 정 전류 방전 실험에서 전류가 방전되는 동안 1초 주기로 개방 회로 전압을 측정함으로써 도 4에 도시된 바와 같은 SOC 와 개방 회로 전압 간의 상호 관계를 나타내는 그래프를 얻었다. First, the parameters of the sliding mode observer

,

,

,

,

,

,

After the secondary battery was fully charged to 4.2V for the purpose of determining and setting the open circuit voltage of the battery for the entire interval (0 to 1) of the SOC, a constant current discharge experiment was performed. In the constant current discharge experiment, a cycle of discharging a current of 5A for 180 seconds and stopping a discharge for 3600 seconds was repeated 20 times as one cycle. For reference, discharging a current of 5A for 180 seconds corresponds to a '1-C rate' of nominal capacity 5.0Ah. In the constant current discharge experiment, when the current of 5A is discharged, the SOC of 5% is reduced. In the constant current discharge experiment, the present inventors obtained a graph showing the correlation between the SOC and the open circuit voltage as shown in FIG. 4 by measuring the open circuit voltage at one second intervals during the current discharge.

,

,

,

및

를 계산하였다. 계산된 결과는

= 17000[F],

= 200 [F],

= 0.001 [Ω] 및

= 0.003 [Ω] 이었고, 전류에 따라 비선형적으로 변화하는

의 경우는 도 5에 도시된 바와 같았다. 이러한 계산 결과와 도 5에 도시된 그래프를 이용하여 슬라이딩 모드 관측기(100)에 포함된 파라미터

,

,

,

,

,

,

를 계산하였다. 그런 다음 계산된 파라미터 값, 정 전류 방전 실험을 통하여 획득한 개방 회로 전압별 SOC 값(도 4 참조) 및 전류에 따른

값을 슬라이딩 모드 관측기의 저장매체에 수록하여 슬라이딩 모드 관측기를 세팅하였다.Then, based on the constant current discharge experiment in the off-line state as described above for the modeling circuit of the secondary battery employed in the present invention

,

,

,

And

Was calculated. The calculated result is

= 17000 [F],

= 200 [F],

= 0.001 [kV] and

= 0.003 [Ω], which varies nonlinearly with current

The case was as shown in FIG. Parameters included in the sliding mode observer 100 using the calculation result and the graph shown in FIG. 5.

,

,

,

,

,

,

Was calculated. Then, the calculated parameter value, SOC value for each open circuit voltage obtained through the constant current discharge experiment (see FIG. 4), and the current

The values were stored in the storage medium of the sliding mode observer to set the sliding mode observer.

를 수집하였다.Once the setting of the sliding mode observer was completed, the sliding mode observer was coupled to the secondary battery in online mode as shown in FIG. 2. At this time, the secondary battery is charged to 4.2V again. Predict the SOC output through the sliding mode observer while replaying the constant current discharge experiment once again in this on-line state.

Was collected.

6 is a graph plotting the error of the discharge current measured in the on-line constant current discharge experiment, the output voltage of the actual battery, the output voltage and the output voltage predicted through the modeling circuit with respect to time t. Referring to the drawings, it can be seen that the predicted value and the actual value of the output voltage are substantially the same except that a slight error occurs at the start and end points of the constant current discharge cycle. On the other hand, an error occurs at the start and end of the constant current discharge cycle because the nominal capacitor C _n has a nonlinear characteristic and the resistance component is changed by SOC in the case of an actual battery.

를 시간 t에 대해 플로팅한 그래프이다. 슬라이딩 모드 관측기의 정 궤환 이득 상수 L₁, L₂ 및 L₃는 각각 0.02, 0.02 및 0.1로 설정하였다. 도 7을 참조하면, 예측된 전지의 출력 전압은 미세한 채터링을 수반하면서 실제 출력 전압을 정확하게 추종하고 있고, 예측된 SOC 또한 방전 초기와 종료 시점을 제외하면 실제 SOC를 정확하게 추종한다는 것을 확인할 수 있다. 방전 초기 및 종료 시점에서 예측된 SOC와 실제 SOC 간에 오차가 발생되는 이유는 불연속적인 전류의 변화와 이에 따른 파급저항 성분 R_p의 급격한 변화 때문이다. 하지만 본 발명에 따른 슬라이딩 모드 관측기는 짧은 시간 안에 SOC예측치가 실제 SOC를 추종하는 우수한 성능이 있음을 확인할 수 있다.7 is an SOC obtained using a sliding mode observer in an on-line state.

Is a graph plotted against time t. The positive feedback gain constants L ₁ , L ₂ and L ₃ of the sliding mode observer were set to 0.02, 0.02 and 0.1, respectively. Referring to FIG. 7, it can be seen that the predicted output voltage of the battery accurately follows the actual output voltage with fine chattering, and the predicted SOC also accurately follows the actual SOC except for the discharge start and end points. . The reason for the error between the predicted SOC and the actual SOC at the beginning and end of the discharge is due to the change in the discontinuous current and the sudden change in the ripple component R _p . However, it can be seen that the sliding mode observer according to the present invention has excellent performance that the SOC prediction value follows the actual SOC in a short time.

Next, the inventors conducted 16 cycles of Urban Dynamometer Driving Schedule (UDDS) experiments in order to investigate the performance of the sliding mode observer according to the present invention in a driving situation of an actual vehicle. At this time, when each UDDS cycle is completed, the discharge pulse of 40A was input to the sliding mode observer for 5 minutes. The reason for entering the discharge pulse is to check the robustness of the sliding mode observer against external noise or disturbance. This UDDS experiment was conducted for 90% to 10% SOC intervals with 5% SOC reduction for each UDDS cycle.

8 is a graph plotting the overall UDDS cycle discharge current obtained in the UDDS experiment, the output voltage predicted through the battery's actual output voltage and the modeling circuit, the error between the actual output voltage and the predicted output voltage, and the SOC's prediction with respect to time t. 9 is an enlarged graph of the UDDS discharge current, the actual output voltage, and the predicted output voltage of one cycle. Referring to the drawings, it can be seen that the error of the output voltage is less than 20mV in the 20% to 80% section of the SOC, and it can be seen that the predicted output voltage follows the actual output voltage fairly accurately even in the actual driving situation.

FIG. 10 is a graph in which the actual and predicted values of SOC obtained for the entire UDDS cycle and the SOC errors are plotted against time t. FIG. 11 is an enlarged view of a portion of FIG. 10. Referring to the drawings, it can be seen that the trajectory of the SOC prediction value follows the actual value of the SOC with fine chattering having a constant upper limit and lower limit, and the average of the predicted value of the SOC is substantially the same as the actual value. . On the other hand, the chattering may be smooth to some extent by replacing the signum function with the saturation function in the design of the sliding mode observer.

As can be seen from the above-described experimental results, the sliding mode observer according to the present invention has a performance of accurately following the actual SOC even though it is designed based on a relatively simple RC circuit modeling. Can be exercised.

As described above, although the present invention has been described by way of limited embodiments and drawings, the present invention is not limited thereto and is intended by those skilled in the art to which the present invention pertains. Of course, various modifications and variations are possible within the scope of equivalents of the claims to be described.

According to an aspect of the present invention, the SOC of the secondary battery is predicted using a sliding mode observer having robustness against disturbance or parameter change. Therefore, the SOC can be accurately predicted even if the constant value of the secondary battery changes or there is a slight error in the measured value of the input current and / or output voltage of the secondary battery.

According to another aspect of the present invention, since a simple RC modeling circuit is used to model the secondary battery, it is possible to simplify the calculation process for predicting the SOC of the secondary battery compared to a technique using a conventional neural network circuit. This allows the implementation of a high performance sliding mode observer against the same hardware specifications.

According to another aspect of the present invention, since the SOC can be predicted within an error range of 3% or less, it may be immediately commercialized in a hybrid electric vehicle.

Claims (17)

Secondary battery output voltage (V _t) induced by RC circuit modeling considering the open circuit voltage (V _oc (Z)) of the secondary battery, the capacitor component (C _p ) due to the polarization effect of the battery, and the resistance component inside the battery. ), The parameters of each state equation for the capacitor across voltage (V _p ) and SOC (Z); A positive feedback gain constant of the observer designed in sliding mode based on each state equation; Open circuit voltage-specific SOC tables; And estimates of V _t , V _p, and Z based on the observer

,

And

A secondary battery state estimation program for calculating a storage medium; Current / voltage detecting means for outputting the sampling time t by measuring the input current (I) and output voltage (V _t) of the secondary battery; And The secondary battery state prediction program, the parameter, the positive feedback gain constant, and the open circuit voltage-specific SOC table are loaded from the storage medium into the memory, and the input current of the secondary battery output from the current / voltage detection means and The output voltage is input, the equation of the observer is established by the loaded parameter and the positive feedback gain constant, and the secondary battery is obtained by using the SOC table for each input current (I), output voltage (V _t ), and open circuit voltage. SOC estimates

SOC prediction apparatus of a secondary battery using a sliding mode observer comprising a; microprocessor for calculating and outputting. The method of claim 1, The resistance component comprises a ripple resistance component R _b , a diffusion resistance component R _p and an ohmic resistance component R _t SOC prediction apparatus of a secondary battery using a sliding mode observer. The method of claim 2, Each state equation for the secondary battery output voltage (V _t ), the voltage across the capacitor (V _p ) and SOC (Z) is represented by the following equation,

The parameters contained in the storage medium are coefficients of the state equation.

,

,

,

,

,

,

as,

And T is a transpose matrix and C _n is a nominal capacitance of a secondary battery. The method of claim 3, The sliding mode observer for the secondary battery output voltage (V _t ), the voltage across the capacitor (V _p ) and SOC (Z) is represented by the following equation,

remind

,

And

Is a positive feedback gain constant stored in the storage medium, and sgn () is a signum function, SOC prediction apparatus of a secondary battery using a sliding mode observer. The method of claim 4, wherein The microprocessor converts the equation of the sliding mode observer into Euler discrete equation and then estimates V _t , V _p and Z by the following equation.

,

And

And calculate

The T _s is a sampling period for the input current (I) and the output voltage (V _t ) of the secondary battery, and K is a secondary battery using the sliding mode observer, characterized in that the sampling is made (K≥1) SOC prediction device. The method of claim 1, The SOC prediction

SOC prediction apparatus for a secondary battery using a sliding mode observer, characterized in that input to a battery management system (BMS). delete The method of claim 1, The storage medium further includes a capacitor component value and a resistance component value included in the RC modeling circuit, SOC prediction apparatus of a secondary battery using a sliding mode observer. The method of claim 8, The storage medium further includes a capacitor component value list and a resistance component value list according to the input current magnitude of the secondary battery. The microprocessor reads from the list a capacitor component value and a resistance component value corresponding to the magnitude of the detected input current at the time t when the input current of the secondary battery is detected, and calculates the parameter in real time. SOC prediction device for a secondary battery using an observer. Secondary battery output voltage (V _t) induced by RC circuit modeling considering the open circuit voltage (V _oc (Z)) of the secondary battery, the capacitor component (C _p ) due to the polarization effect of the battery, and the resistance component inside the battery. ), The parameters of each state equation for the capacitor across voltage (V _p ) and SOC (Z); A positive feedback gain constant of the observer designed in sliding mode based on each state equation; Open circuit voltage-specific SOC tables; And estimates of V _t , V _p, and Z based on the observer

,

And

In the method for predicting the SOC of a secondary battery using a storage medium containing a secondary battery state prediction program for calculating a; (a) loading an SOC table for each program, parameter, positive feedback gain constant, and open circuit voltage from the storage medium into a memory; (b) receiving the input current I and the output voltage V _t of the secondary battery from the current / voltage detection means at the sampling time t; And (c) The equation of the observer is established by the loaded parameters and the positive feedback gain constant, and the SOC prediction value of the secondary battery using the SOC table for each input current (I) and output voltage (V _t ) and open circuit voltage.

Computing and outputting; SOC prediction method of a secondary battery using a sliding mode observer comprising a. The method of claim 10, The resistance component comprises a ripple resistance component R _b , a diffusion resistance component R _p and an ohmic resistance component R _t SOC prediction method of a secondary battery using a sliding mode observer. The method of claim 11, Each state equation for the secondary battery output voltage (V _t ), the voltage across the capacitor (V _p ) and SOC (Z) is represented by the following equation,

The parameter contained in the storage medium is a coefficient of the state equation.

,

,

,

,

,

,

as,

And T is a transpose matrix and C _n is a nominal capacitance of a secondary battery. The method of claim 12, The sliding mode observer for the secondary battery output voltage (V _t ), the voltage across the capacitor (V _p ) and SOC (Z) is represented by the following equation,

remind

,

And

Is a positive feedback gain constant stored in the storage medium, and sgn () is a signum function, SOC prediction method of a secondary battery using a sliding mode observer. The method of claim 13, wherein step (c) comprises: (c1) Estimates of the output voltage (V _t ), the voltage across the capacitor (V _p ), and the SOC (Z) of the secondary battery by substituting the parameters and positive feedback gain constants stored in the storage medium into the following equation.

,

And

Establishing a mathematical algorithm for the calculation of; And

(c2) The secondary battery is substituted with the input current (I) and output voltage (V _t ) input at the sampling time t, and the open circuit voltage and SOC at the time t contained in the SOC table for each open circuit voltage in the discrete equation. Output voltage, the voltage across the capacitor and the estimated SOC

,

And

Comprising: wherein, T _s is a sampling period for the input current (I) and output voltage (V _t ) of the secondary battery, K is characterized in that the sampling is made (K≥1) SOC prediction method of a secondary battery using a sliding mode observer. The method of claim 10, wherein in step (c), The calculated SOC prediction value is input to a battery management system (BMS) SOC prediction method of a secondary battery using a sliding mode observer, characterized in that. delete The method of claim 10, The storage medium further includes a capacitor component value list and a resistance component value list according to the input current magnitude of the secondary battery. The step (c) may include: calculating the parameter in real time by reading a capacitor component value and a resistance component value corresponding to the magnitude of the detected input current at the time t when the input current of the secondary battery is detected; Including more, In the step (c), the parameter used in the establishment of the mathematical algorithm is a SOC prediction method of a secondary battery using a sliding mode observer, characterized in that the parameter calculated in real time.

Images