WO2019066278A1

WO2019066278A1 - Method for measuring entropy of battery using kalman filter

Info

Publication number: WO2019066278A1
Application number: PCT/KR2018/010233
Authority: WO
Inventors: 이상국; 기욤테네지; 한석균
Original assignee: 한국과학기술원
Priority date: 2017-09-29
Filing date: 2018-09-03
Publication date: 2019-04-04
Also published as: KR101946784B1

Abstract

According to a method of the present invention, a Kalman filter unit comprises: a first Kalman filter that performs a filter operation in accordance with changes in the state of charge of a battery; and a second Kalman filter that performs a filter operation according to changes in the number of uses of the battery, wherein the first Kalman filter, in a state where the first Kalman filter has received a plurality of entropy values including an actual entropy value and an expected entropy value, on the basis of changing state of charge of the battery, outputs a corrected actual entropy change rate, at the same time that the first Kalman filter is repeatedly operated, and the second Kalman filter periodically updates an entropy model, in a state where the second Kalman filter has received both the actual entropy change rate of the first Kalman filter and an expected aging entropy change rate reflecting aging due to an increase in the number of uses of the battery, and the first Kalman filter tracks an entropy change by means of information about changes in the state of charge of the battery supplied from the entropy model.

Description

How to measure the entropy of a battery using a Kalman filter

The present invention relates to a technique for periodically updating an entropy model used in a Kalman filter while a Kalman filter is operating to extract entropy in real time from a battery management system (BMS). More specifically, (OCV), the Kalman filter is introduced under real-time measurement of the open-circuit voltage. By correcting the dispersion of the entropy change rate due to the measurement dispersion of the open-circuit voltage in the equilibrium state, And a method for measuring the rate of change.

Lithium-ion batteries are currently being used as energy sources for various applications due to their high output and high energy characteristics, and their application range is expected to be broader. However, battery capacity reduction and safety issues are a challenge to be solved in lithium ion batteries.

However, it has been found that the degree of entropy change of the battery can be utilized as an indicator of the capacity reduction and safety of the battery, and researches thereof are actively under way. That is, since the entropy profile of the battery has a characteristic that it changes with use of the battery, if the above-mentioned differential entropy is known, it is advantageous that it can be used as an index for predicting the aging state and the risk of the battery do.

Electrochemical Thermodynamics Measurement (ETM) is a method of measuring the entropy change of a battery.

The ETM provides a measure of the amount of change in open-circuit voltage (OCV) when the battery temperature is forced to change while the battery is in equilibrium.

The entropy of the battery is not defined as a single value but is measured separately for the state of charge (SoC) of the battery. The entropy profile according to the SoC is measured by measuring the desired step within 0% to 100%.

The ETM measures the change of the open-circuit voltage (OCV) of the battery with temperature, which is measured by ETMS. However, there is a problem that the use of ETMS is limited to non-commercial equipment.

The measurement of entropy change using ETMS generally proceeds as follows.

First, the OCV of the chemical equilibrium state is measured through relaxation at a battery temperature of 25 ° C and 0% SoC. Then, change the temperature of the battery to 20 ° C, 15 ° C, and 10 ° C, measure the OCV at each temperature, and measure the tilt of OCV according to the temperature. This process is performed for each SoC step, and as a result, the entropy change in the entire SoC is measured.

However, the ETMS is difficult to use as a non-commercial equipment, and the temperature changing device is specially manufactured to insert a specific battery into the device, thereby limiting the number of the battery models and the number of the batteries that can be measured at one time.

In addition, it takes a long time to reach the chemical equilibrium state, which causes a long measurement time.

On the other hand, in order to measure the entropy change of the battery, it is premised that the battery measures the change in the OCV with respect to the measured voltage (OCV) and the temperature change in the electrical / chemical / thermal equilibrium state.

However, since the time required for the battery to reach a thermal equilibrium state is long for about 24 hours, and a dynamically programmable dedicated temperature changing device is required, the battery has a limitation that is difficult to apply to BMS There is no such case applied to BMS so far.

An entropy model used in a Kalman filter is periodically updated by operating a Kalman filter to extract entropy in real time from a BMS. In an active battery, an open-circuit voltage OCV), the Kalman filter is introduced under real-time measurement of the open-circuit voltage, and the entropy change rate due to the measurement dispersion of the equilibrium open-circuit voltage is corrected to measure the entropy change rate in real time without reaching the equilibrium state And the like.

According to an aspect of the present invention, there is provided a method for measuring entropy of a battery using a Kalman filter,

Wherein the Kalman filter unit includes: a first Kalman filter that operates according to a change in a state of charge (SoC) of the battery; And a second Kalman filter for performing a filter operation in accordance with a change in the number of times of use of the battery,

Wherein the first Kalman filter outputs a corrected actual entropy change rate in a state in which a plurality of entropy values including an actual entropy value and an estimated entropy value are input through the changed battery charging state and is repeatedly operated, The filter receives the actual entropy change rate of the first Kalman filter and receives the estimated change rate of the aging entropy reflecting the aging due to the increase in the battery usage count, outputs the estimated change rate of the aging entropy, and updates the entropy model periodically , And the first Kalman filter tracks the entropy change by the battery charge state change information supplied from the entropy model updated through the operation of the second Kalman filter.

According to another aspect of the present invention, there is provided a system for measuring entropy of a battery using a Kalman filter, comprising: a voltage measuring unit; A temperature measuring unit 20; An entropy calculation unit 30 for calculating entropy based on temperature and voltage information obtained through the voltage measurement unit 10 and the temperature measurement unit 20; An SOC checking unit 42 for checking the battery charging state in real time; A first modeling unit 40 for outputting the enthalpy change rate and the expected entropy change rate; A second modeling unit 50 that performs modeling with a rate of change according to the number of times of use of the battery in a state where a change occurs at a specific charging point of the battery as the battery ages; And a Kalman filter unit 60 for providing a corrected entropy change rate through data exchange between the entropy calculation unit 30 and the first and

second modeling units

40 and 50. The Kalman filter unit 60 Includes: a first Kalman filter that operates according to a change in a state of charge (SoC) of the battery; And a second Kalman filter that operates according to the change in the number of times of battery cycle use.

The first Kalman filter 62 inputs two entropy predictive values such as the actual entropy value? Smesas supplied from the entropy calculation unit 30 and the estimated entropy value? Spred supplied from the first modeling unit 40 The actual entropy is output, the SoC is changed while the charge / discharge process is performed, and the Kalman filter is repeatedly operated.

The second Kalman filter 64 is provided with a predicted aging entropy change rate (ΔSaged_pred) that reflects a change in aging of the battery from the second modeling unit 50, And an aging entropy change rate (? Saged_meas) as a result of the measurement.

The estimated aging entropy change rate (ΔSaged_est) information is supplied to the first modeling unit (40) while the second Kalman filter (64) is operated. In the first modeling unit (40), the enthalpy change rate And the entropy change rate is obtained.

The present invention provides a method of measuring entropy of a battery in real time using the Kalman filter according to the present invention. In order to obtain the OCV slope with respect to the temperature, a temperature control device is required and a chemical equilibrium state is reached. This overcome the problem that it is difficult to implement in BMS due to the practical necessity.

The present invention eliminates the need for a temperature changing device because the entropy change rate can be easily obtained by measuring the equilibrium state OCV and the temperature when the enthalpy change is known in advance. That is, the enthalpy change is advantageous in that it can use pre-measured data in the laboratory because there is little variation among individuals in the same kind of battery and the change does not occur even when the battery ages.

In the present invention, a technique for estimating the equilibrium state OCV on an operating battery is already available, and by measuring OCV in real time, it is possible to easily obtain the entropy change rate in real time without reaching an equilibrium state.

There has been a problem of lowering accuracy in the conventional technique of estimating the equilibrium state OCV. The present invention introduces a Kalman filter to correct the entropy change rate dispersion caused by the equilibrium OCV measurement dispersion.

1 is a diagram illustrating an entropy extraction system using a dual Kalman filter according to an embodiment of the present invention.

FIG. 2 is a graph showing that the Kalman filter is repeatedly operated according to a certain period of time while the SoC of the battery is changed, and gradually converges to the actual value as the number of operations increases.

FIG. 3 shows a graph in which a change occurs at a specific SoC point according to aging of the battery and is modeled by a rate of change according to the number of times of use.

Hereinafter, embodiments of the present invention will be described in detail with reference to the accompanying drawings. It should be understood, however, that the invention is not limited to the disclosed embodiments, but is capable of other various forms of implementation, and that these embodiments are provided so that this disclosure will be thorough and complete, It is provided to let you know completely. Wherein like reference numerals refer to like elements throughout.

It should be noted that, in adding reference numerals to the constituent elements of the drawings, the same constituent elements are denoted by the same reference numerals even though they are shown in different drawings. In the following description of the present invention, a detailed description of known functions and configurations incorporated herein will be omitted when it may make the subject matter of the present invention rather unclear.

The thermodynamic terms of a battery used in connection with the entropy measurement in the present invention are disclosed in Korean Patent Laid-Open Nos. 10-2017-0093482 and 10-2017-0059208 filed by the present applicant.

For convenience, the terms used in this specification are described below.

An open circuit voltage (OCV) is a voltage between two positive and negative electrodes when no load is applied to the battery, that is, when no current is externally discharged. The maximum value of the open- It becomes equal to the value of the electromotive force.

The State of Charge (SoC) represents the state of charge and is equivalent to the fuel gauge of the battery, in units of percentage points. Specifically, 0% indicates fullness and 100% indicates fullness. The SoC is mainly used to indicate the current state of charge of the battery in use.

A battery management system (BMS) is an electronic system that manages a rechargeable battery (cell or battery pack), which protects the battery from operating outside the safe operating area, monitors the status of the battery, (Not limited to this task) of computing, reporting, reporting its data, controlling its environment, authenticating and / or balancing the battery.

The enthalpy is equal to the total calorific value of the system, equal to the internal energy of the system plus the products of pressure and volume. The change in enthalpy of the system is associated with a particular chemical process.

Entropy is a thermodynamic quantity (state function) that represents the thermal energy of a system that is not useful for converting to mechanical work, and is often interpreted as the degree of randomness or randomness of the system.

The present invention focuses on overcoming the technical limitations in applying electrochemical thermodynamic measurements (ETM) to a battery management system (BMS).

Specifically, the entropy of a battery is presumed to measure the level of change of the OCV with respect to the open-circuit voltage and the temperature change, which is the voltage measured in the electric / chemical / thermal equilibrium state of the battery.

However, since the time required for the battery to reach a thermal equilibrium state is long for about 24 hours, and a dynamically programmable dedicated temperature changing device is required, the battery has a limitation that is difficult to apply to BMS There is no case applied to the BMS until now, and it is intended to improve the entropy change rate measurement by using the correction system using the Kalman filter.

First, we explain the ETM theory.

The second law of thermodynamics is expressed by the following equation (1).

. . . (One)

(G) is the total energy of the battery, G is the available energy of the battery (ΔG = -nF × OCV), TxΔS (temperature × Entropy) )

When the battery is in an electrochemical / thermal equilibrium state, applying the Ellingham approximation, it can be assumed that ΔS and ΔH are not a function of temperature over a certain temperature range.

When the above equation (1) is differentiated by temperature, entropy is defined by the following equation (2).

.(2)

(Where n represents the exchange capacity of electrons in a typical basic reaction, and F is a Faraday constant).

As shown in Equation (2), a temperature controller is required to obtain the OCV slope with respect to temperature, and it takes a long time to reach the chemical equilibrium state, which makes it difficult to implement in the BMS.

Accordingly, the present invention eliminates the necessity of the temperature changing device through the following equation (3) since the entropy change rate can be easily obtained by measuring the equilibrium state OCV and temperature when the enthalpy change is known in advance.

. (3)

Hereinafter, an entropy extraction system using a dual Kalman filter will be described with reference to FIG.

The entropy extraction system includes an OCV acquisition unit 12 for extracting OCV based on data measured by the voltage measurement unit 10 and the voltage measurement unit 10, a temperature measurement unit 20, a voltage measurement unit 10, An entropy calculation unit 30 for calculating an entropy based on the temperature and voltage information acquired through the unit 20, an SOC checking unit 42 for checking the battery charging status in real time, an initial value input unit 44, A first modeling unit 40 for outputting an enthalpy change rate and an expected entropy change rate based on the information of the unit 42 and the initial value input unit 44, A dual Kalman filter unit 60 for providing a corrected entropy change rate through data exchange between the entropy calculation unit 30 and the first and

second modeling units

40 and 50, .

The dual Kalman filter unit 60 includes a first Kalman filter 62 that operates according to the SoC change and a second Kalman filter 64 that operates according to the change in the number of times the battery is used.

The first modeling unit 40 measures the entropy profile of the battery in a laboratory environment and models the model as a function according to the SoC. That is, through the modeling information linearized for each section through the charge state information of the battery, which is checked in real time by the SOC checking unit 42, the initial enthalpy change rate and the initial estimated entropy change rate, which are confirmed by the initial value input unit 44, The predicted enthalpy change rate and the estimated entropy change rate (ΔSpred) are obtained in the SOC section of the battery. The predicted enthalpy change rate and the estimated entropy change rate are provided to the entropy calculation unit 30 and the dual Kalman filter unit 60, respectively.

In the present invention, by measuring the real time OCV and the temperature (T) using the known known equilibrium state OCV prediction technique and substituting it into the equation (3), the measured entropy change rate can be obtained.

The first Kalman filter 62 receives the two entropy predicted values such as the actual entropy value? Smesas supplied from the entropy calculation unit 30 and the estimated entropy value? Spred supplied from the first modeling unit 40, As the entropy is output, the SoC is changed and the Kalman filter is repeatedly operated as it is charged and discharged.

FIG. 2 shows that the Kalman filter is repeatedly operated according to a constant period as the SoC of the battery is changed, and gradually converges to the actual value as the number of operations increases. Specifically, the axis of abscissa shows the variation axis of the SoC as the number of times of battery usage increases, and the axis of ordinate shows the change rate of entropy. At the beginning of the operation of the battery, although the difference between the output value of the Kalman filter and the actual entropy value is large, it can be confirmed that the error is reduced through continuous repetitive operation.

On the other hand, entropy changes through the process of increasing the number of times of use while using the battery.

Therefore, there is a need for a task of modeling the entropy that changes due to the aging of the battery in advance in the laboratory, and estimating the entropy according to the aging of the battery.

As shown in FIG. 3, in the second modeling unit 50, a change occurs at a specific SoC point according to aging of the battery, and is modeled by a rate of change according to the number of times of use. That is, the entropy profile is measured separately according to the state of charge (SoC) of the battery formed along the horizontal axis, and the entropy profile according to the SoC is measured by measuring the desired step within 0% to 100%. In the above, it reflects the degree of change at a specific SoC point as the battery ages.

The predicted aging entropy change rate (ΔSaged_pred) reflecting the change with aging of the battery is supplied from the second modeling unit 50 to the second Kalman filter 64 and the corrected result of the measurement derived from the first Kalman filter 62 The aging entropy change rate ([Delta] Saged_meas) is input to the measured value of the second Kalman filter 64.

The process of measuring the entropy of the battery according to an embodiment of the present invention does not consider the aging when the SoC is changed in the same cycle so that the first modeling unit through the second Kalman filter 64 40 is operated every one cycle or a plurality of cycles. Therefore, at the calculation of entropy at one point, the first Kalman filter 62 functions sufficiently to operate once without considering the aging through the second Kalman filter 64.

Meanwhile, it is also possible that the first and second Kalman filters 62 and 64 are operated and updated at every measurement time when the SoC is changed in the same cycle.

The estimated aging entropy change rate (ΔSaged_est) information as the second Kalman filter 64 is operated is supplied to the first modeling unit 40. In the first modeling unit 40, the enthalpy change rate and the estimated entropy change rate Perform the process of obtaining again. The first Kalman filter 62 receives the two entropy predicted values, such as the actual entropy value supplied from the entropy calculator 30 and the predicted entropy value supplied to the first modeling unit 40, and outputs the actual entropy, The SoC is changed and the Kalman filter is repeatedly operated.

Accordingly, the first and second Kalman filters 62 and 64 of the dual Kalman filter unit 60 periodically update the entropy model through the first and

second modeling units

40 and 50, Successfully track changes.

The present invention provides a method of detecting a true value by correcting an error of a measured value due to inaccuracy or other reasons of an existing measurement sensor using a dual Kalman filter.

In particular, the method of directly measuring a desired value in a process of using a dual Kalman filter, and the method of predicting through a model by modeling a system to be measured are used in an overlapping manner.

Although the direct measurement method generally has a large measurement error, the error in each measurement is maintained at the same level.

The modeling prediction method is to predict and model the system to be measured. If the modeling is accurate, the prediction error is small, but when the value is continuously predicted through the modeling, the error is increased as the previously predicted error continues to accumulate By using the dual Kalman filter, the measurement error is minimized by periodically receiving the two values of the direct measurement method and the modeling prediction method.

In the process of using the dual Kalman filter, although the error may be initially large after the operation of the Kalman filter, the iteration is repeated over time, and finally the measurement value having the minimum error is calculated.

As described above, according to the method of measuring entropy of a battery using a dual Kalman filter according to the present invention, a temperature control device is required to obtain the OCV slope with respect to the temperature, and a chemical equilibrium state is reached at the same time. This overcome the problem that it is difficult to implement in BMS due to the practical necessity

The foregoing description is merely illustrative of the technical idea of the present invention, and various changes and modifications may be made by those skilled in the art without departing from the essential characteristics of the present invention. Therefore, the embodiments disclosed in the present invention are intended to illustrate rather than limit the scope of the present invention, and the scope of the technical idea of the present invention is not limited by these embodiments. The scope of protection of the present invention should be construed according to the following claims, and all technical ideas within the scope of equivalents should be construed as falling within the scope of the present invention.

Claims

A method for measuring entropy of a battery using a Kalman filter unit,

The Kalman filter unit may be configured to change the state of charge (SoC)

A first Kalman filter for filtering; And

And a second Kalman filter for performing a filter operation in accordance with a change in the number of times of use of the battery,

In the first Kalman filter, an actual entropy

Value and a predicted entropy value, and outputs the corrected actual entropy change rate while being repeatedly operated,

The second Kalman filter receives the actual entropy change rate of the first Kalman filter and the estimated aging entropy change rate reflecting the aging due to the increase in the battery usage count and outputs the estimated aging entropy change rate to output the entropy model Periodically update,

Wherein the first Kalman filter tracks an entropy change by battery charge state change information supplied from an entropy model updated through operation of the second Kalman filter,

A method of measuring the entropy of a battery using a Kalman filter.
A voltage measuring unit;

A temperature measuring unit;

Based on the temperature and voltage information obtained through the voltage measurement unit and the temperature measurement unit

An entropy calculation unit for calculating entropy;

A SOC checking unit for checking the battery charging state in real time;

A first modeling unit for outputting an enthalpy change rate and an expected entropy change rate;

Depending on the aging of the battery, a change occurs at a specific charge point of the battery

A second modeling unit that performs modeling with a rate of change according to the number of times the battery is used; And

Wherein the entropy calculation unit and the first and second modeling units exchange data,

And a Kalman filter unit for providing a rate of change of entropy,

Wherein the Kalman filter unit includes: a first Kalman filter that operates according to a change in a state of charge (SoC) of the battery; And

And a second Kalman filter for performing a filter operation in accordance with a change in the number of times of battery cycle use

doing,

Entropy Measurement System of Battery Using Kalman Filter.
3. The method of claim 2,

In the first Kalman filter, the actual entropy supplied from the entropy calculation unit

And the predicted entropy value (? Spred) supplied from the first modeling unit, and outputs the actual entropy, and the SoC is changed while the charge / discharge process is performed.

Entropy Measurement System of Battery Using Kalman Filter.
3. The method of claim 2,

The second Kalman filter is a predicted aging entity that reflects a change in aging of the battery

Wherein the second modeling unit receives the rate of change of the measured entropy (ΔSaged_pred) and receives the aging entropy change rate (ΔSaged_meas), which is a corrected measurement result from the first Kalman filter,

Entropy Measurement System of Battery Using Kalman Filter.
5. The method of claim 4,

The aging entropy change rate (ΔSaged_est) estimated as the second Kalman filter is operated is supplied to the first modeling unit, and the first modeling unit calculates a change rate of the enthalpy and a predicted entropy change rate at a specific SoC point.

Entropy Measurement System of Battery Using Kalman Filter.