KR20190080102A - A method of estimating state of charge of battery and an apparatus for managing of battery - Google Patents

Abstract

The present invention relates to a method for estimating a state of charge of a battery and a battery management device. According to an embodiment of the present invention, the method comprises the steps of: determining whether the battery mounted on a device using the battery is first used; performing initialization when the battery is first used; performing a long term energy process when performing the initialization; predicting a charge state and a resistance value of the battery through the long term energy process; performing a short term energy process when not performing the initialization; determining whether to enter an idle state based on a current and a voltage value of the battery; re-adjusting the charge state value of the battery by acquiring an open circuit voltage value of the battery when entering the idle state; and continuously performing the short term energy process to predict a current resistance and a current open circuit voltage value of the battery when not entering the idle state. Thus, the clarity of charge state estimation of the battery can be improved.

Description

BACKGROUND OF THE INVENTION 1. Field of the Invention [0001] The present invention relates to a method of estimating a state of charge of a battery,

The present invention relates to a method of estimating a state of charge of a battery and a battery management apparatus, and more particularly, to a method of estimating a state of charge of a battery by various methods depending on whether the battery is used or not, .

Similar to increasing demand for performance from small mobile devices and Internet (IoT) devices, the battery management system, which is used to secure stability and guarantee the available time of the battery, BMS) has been growing. Obviously, a battery observation technique for accurately measuring the charge state at low cost may be essential for efficient power management in small mobile or object Internet devices with limited hardware resources and power supply.

Commonly used techniques for estimating the state of charge (SOC) for handheld devices can be broadly divided into open-circuit voltage (OCV) -based estimation methods and Coulomb counting-based estimation methods have. The open-circuit voltage based method can measure the voltage of the battery in a resting state in which the apparatus is not used, and can bring the terminal voltage Vt close to the open-circuit voltage value. However, for a dormant battery, the devices must maintain a very low current discharge condition for a period of extended time. Therefore, it can be difficult for the devices to continuously estimate the state of charge.

In addition, the coulomb counting-based method can obtain the charge amount by scaling the discharge current for a specific period, and can estimate the value of the charge state. However, if the initial charge state value is not known, the method can not clearly know the current charge state level. And, errors can be cumulatively increased as the coulomb counting process continues due to errors in the current measurement. Techniques using a combination of the two methods can improve the clarity of the charge state measurement. However, because small handheld or point-to-point Internet devices have low power supply sources, estimating the state of charge by simultaneously measuring voltage and current can increase the cost of the associated hardware.

As another method, there is a low-cost method of estimating the state of charge by intermittently measuring only the terminal voltage. The above method uses only the terminal voltage by using the similarity between the difference between the terminal voltage and the open-circuit voltage in a specific section, so that the cost can be reduced. In addition, since the method does not require the cost for additional circuitry associated with current measurement, it is possible to ascertain the approximate charge state estimation probability.

However, such a method may reduce the clarity of the charge state estimation in a variable discharge period in which the discharge current temporarily changes at a time. Due to these limitations, the method can cause errors in the charge state measurement substantially in recent portable or object Internet devices where the computational performance is improved and the power usage pattern is highly variable.

SUMMARY OF THE INVENTION The present invention provides a method of estimating a charged state of a battery by various methods depending on the state of the battery.

Another aspect of the present invention is to provide a battery management apparatus that performs a method of estimating a state of charge of a battery that minimizes the influence of errors accumulated when estimating a state of charge of the battery.

According to an embodiment of the present invention, there is provided a method for controlling a battery, the method comprising: determining whether the battery mounted on a device using the battery is used for the first time; Performing the initialization if the battery is first used; Performing the long-term energy process when performing the initialization; Estimating a charged state value and a resistance value of the battery through the long-term energy process; Performing the short-time energy process if the initialization is not performed; Determining whether to introduce the battery into a dormant state based on a current value and a voltage value of the battery; Acquiring an open-circuit voltage value of the battery to readjust the state of charge of the battery when the battery is introduced into the dormant state; And continuing the short-time energy process to predict the current resistance value and the current open-circuit voltage value of the battery if the battery is not introduced into the dormant state. The step of performing the initialization may set an initial terminal voltage, an initial charge state value, and an initial resistance value.

In one embodiment, performing the long-term energy process comprises: estimating an open-circuit voltage value using the initial terminal voltage value and the measured terminal voltage value; Measuring a charge state difference value using the open-circuit voltage value, the initial charge state value, and the measured charge state value; Estimating an average current value using the charge state difference value; Estimating a charge state value of the battery by obtaining the open-circuit voltage value using a resistance value obtained based on the average current value and the initial resistance value; And estimate the charge state value and the resistance value of the long-term energy based on the predicted open-circuit voltage value and the obtained resistance value.

The step of performing the short-time energy process may include estimating an open-circuit voltage, a charge state value, and a resistance value of the battery; Estimating a current value using the open-circuit voltage, the charge state value, and the resistance value; And estimating a current charging state value and a current resistance value of the battery using the current value and the difference value of the charging state value.

The step of determining whether to introduce the dormant state may be introduced to the dormant state when the current value of the battery is less than the minimum current value or the period in which the current terminal voltage value is kept constant is longer than the threshold time. The step of resetting the open-circuit voltage value of the battery includes performing at least one of the long-time energy process and the short-time energy process again.

According to another embodiment of the present invention, a method for estimating the state of charge of the battery may be a computer-readable recording medium recording a program for executing the method in a computer.

According to another embodiment of the present invention, in a battery management apparatus, the battery management apparatus includes a processor, and the processor is configured to: determine whether the battery mounted in the apparatus using the battery is used for the first time; Performing an initialization if the battery is first used; Performing the longtomes energy process when performing the initialization; Predicting a charged state value and a resistance value of the battery through the long-term energy process; Performing the short-time energy process when the initialization is not performed; Determining whether to introduce into a dormant state based on a current value and a voltage value of the battery; Acquiring an open-circuit voltage value of the battery to re-predict the state of charge of the battery when the battery is introduced into the dormant state; And if it is not introduced into the dormant state, performing the operation of continuing the short-time energy process to predict the current resistance value and the current open-circuit voltage value of the battery.

The operation of performing the initialization may set an initial terminal voltage, an initial charge state value, and an initial resistance value.

In one embodiment, the operation of performing the long-term energy process comprises: predicting an open-circuit voltage value using the initial terminal voltage value and the measured terminal voltage value; Measuring the charge state difference value using the open-circuit voltage value, the initial charge state value, and the measured charge state value; Predicting an average current value using the charge state difference value; Predicting the open-circuit voltage value using the average current value and the resistance value measured based on the initial resistance value; And estimating a charge state value and a resistance value of the long-term energy based on the measured open-circuit voltage value and the measured resistance value.

Wherein the operation of performing the short-term energy process comprises: predicting an open-circuit voltage, a charge state value, and a resistance value of the battery; Estimating a current value using the open-circuit voltage, the charge state value, and the resistance value; And a current value of the battery and a current resistance value using the difference between the current value and the charge state value.

The operation of determining whether the battery is introduced into the dormant state can be introduced to the dormant state when the current value of the battery is smaller than the minimum current value or the period in which the current terminal voltage value is kept constant is longer than the threshold time. The operation of obtaining the open-circuit voltage value of the battery and re-predicting the state of charge of the battery may include performing at least one of the long-time energy process and the short-time energy process again.

According to an embodiment of the present invention, it is possible to estimate the state of charge of the battery by various methods according to the state of the battery, to estimate the state of charge of the battery which can minimize the added hardware while improving the clarity of the state of charge estimation of the battery, To provide a predictable method.

According to another embodiment of the present invention, there is provided a battery management apparatus that performs a method of estimating or predicting a state of charge of a battery that minimizes the influence of errors accumulated when estimating a state of charge of a battery by applying a provisional correction method . It is another object of the present invention to provide a method for estimating a state of charge of a battery at low cost and low power by measuring only the terminal voltage and using the terminal voltage.

FIG. 1 illustrates a configuration of a battery charging state estimating apparatus according to an embodiment of the present invention.
2 is a flowchart illustrating a long-term energy estimation process according to an embodiment of the present invention.
3 is a flowchart illustrating a process of estimating a short-term energy according to an embodiment of the present invention.
4 is a flowchart illustrating a method of estimating a state of charge of a battery according to an embodiment of the present invention.
5 is a flowchart illustrating a detailed method for estimating a state of charge of a battery according to an embodiment of the present invention.
6A and 6B are graphs of experimental results for estimating initial charge state values in a stable discharge state.
FIGS. 7A and 7B are graphs of experimental results obtained by estimating initial charge state values in a variable discharge state.
8A to 8D are graphs of experimental results that estimate the state of charge in a stable discharge scenario.
FIGS. 9A to 9D are graphs of experimental results that estimate the state of charge in a variable discharge scenario.
FIGS. 10A to 10D are graphs of experimental results that estimate the state of charge in a variable discharge scenario.

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

The embodiments of the present invention are described in order to more fully explain the present invention to those skilled in the art, and the following embodiments may be modified into various other forms, It is not limited to the embodiment. Rather, these embodiments are provided so that this disclosure will be more faithful and complete, and will fully convey the scope of the invention to those skilled in the art.

Like numbers refer to like elements in the drawings. Also, as used herein, the term "and / or" includes any and all combinations of any of the listed items.

The terms used herein are used to illustrate the embodiments and are not intended to limit the scope of the invention. Also, although described in the singular, unless the context clearly indicates a singular form, the singular forms may include plural forms. Also, the terms "comprise" and / or "comprising" used herein should be interpreted as referring to the presence of stated shapes, numbers, steps, operations, elements, elements and / And does not exclude the presence or addition of other features, numbers, operations, elements, elements, and / or groups.

Hereinafter, embodiments will be described in detail with reference to the accompanying drawings. In the following description, 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 image processing method according to an exemplary embodiment may be performed by an image processing apparatus. Like reference symbols in the drawings denote like elements.

In the method of estimating the state of charge of the battery of the present invention, a temporary state based charge state estimation method and a provisional interim correction method based on a long-term (OCV, resistance [R]) are used. This is to improve the clarity of charge state estimation within a variety of variable discharge states for low cost charge state estimation that only measures terminal voltages. The method of estimating the state of charge of the battery according to an embodiment of the present invention will be described later in comparison with other existing methods in the variable discharge test that the proposed method shows an estimation error of the state of charge state value.

A method for estimating a state of charge in portable devices that typically use only terminal voltage without a current sensor, the method comprising: modeling a battery in a primary RC circuit and determining a characteristic of a capacitor voltage change and a register voltage change toward a terminal voltage in a discharge state, And a current is measured based on a low frequency band filter (LPF), and a coulomb count is performed. The method can estimate the state of charge by a simple process as compared with other methods, but can not charge the initial state of charge, and when the discharge current changes rapidly, the accuracy of the estimated state of charge state value can be lowered have.

FIG. 1 illustrates a configuration of a battery charging state estimating apparatus according to an embodiment of the present invention.

Referring to FIG. 1, a battery management system (BMS) 102 may estimate the state of charge of the battery 104 in real time according to one embodiment. The battery management device 102 can use the observable state information of the battery 104 measured through the sensor to estimate the internal state of the battery 104 or an unobservable variable in real time. The battery management device 102 can accurately estimate, track, or predict the unobservable variables of the battery 104 by assimilation of the measured state information with the estimated charge state in advance.

Here, the observable state information may include one or more state information of the voltage, current, and temperature of the battery 104. Unobservable variables may include State of Charge (SoC) or State of Health (SoH). The state of charge of the battery 104 may include at least one of terminal voltage of the battery 104, terminal current, charge / discharge efficiency, internal temperature, charge rate, and charge density.

The battery management apparatus 102 may be configured to estimate the battery voltage of the battery 104 using a short-term energy (STE) estimation method, a long-term energy (LTE) estimation method, and a provisional correction method. Terminal voltage, charge state value, resistance value, and the like can be obtained. In detail, the state of charge of the battery 104 is indicated through a short-term energy estimation method or a long-term energy estimation method, depending on whether the battery 104 to the device on which the battery 104 is mounted has been previously used. Value can be obtained.

The battery charge state estimation apparatus 100 may include a battery management device 102 and a battery 104. [ In one embodiment, the battery management device 102 may be external to the battery 104 separately. The battery management device 102 may be included in an electronic device such as a mobile phone or a laptop or an electric vehicle. A method of estimating the state of charge of the battery 104 may be implemented in the battery management apparatus 102. [ The method of estimating the state of charge of the battery 104 may allow the battery management apparatus 102 incorporated in the battery state estimation apparatus 100 to perform functions for estimating the state of charge of the battery 104. [ The battery 104 may include, for example, a lithium-ion cell.

The battery management device 102 may be applied to intelligent battery management required in an electronic device or an electric vehicle or the like. The battery management device 102 may react intelligently to the behavior of a user, driver, or pilot in different environments. For example, the battery management device 102 can help the autonomous drive controller of the unmanned electric vehicle make the appropriate decisions when there is uncertainty about the state of the battery 104 of the unmanned electric vehicle. The battery management device 102 may estimate the unobservable variables of the battery 104 with a certain level of reliability or more.

For example, an electric vehicle satisfies a high output by using a battery pack composed of lithium-ion cells arranged in a serial and / or parallel structure in general. The battery management device 102 may use a sensor to monitor a lithium-ion cell or to estimate observable parameters such as temperature, current or voltage. The sensor may be an on-board sensor and may be disposed within the battery 104, and the battery management device 102 may also be disposed on-board.

The battery management apparatus 102 can estimate unobservable variables and internal states in real time by assimilating data in real time. The battery management device 102 can ensure safe operation by continuously monitoring each lithium-ion cell and estimating unobservable variables of the battery 104, such as a state of charge or a state of deterioration. As such, the battery management apparatus 102 can solve the new safety problem arising from the scaled-up battery pack configuration.

The battery management device 102 may provide a framework for obtaining parameter information or status information of the battery 104, such as current, voltage, or temperature measured by the sensor. The battery management device 102 may compare the acquired information with an initially estimated charge state and estimate the current state of charge of the battery 104 based on the comparison result.

The battery management device 102 can estimate in real time information about the unobservable and internal states of the battery 104 by measuring the observable state of the battery 104, such as voltage, current or temperature, from the sensor. The battery management device 102 performs computationally efficient data assimilation where the estimated state of charge can be used to estimate the current state of charge of the battery 104 by being combined with information of the observable parameters of the battery 104. [ The battery management device 102 can be modified for real-time implementation and is computationally economical. The battery management device 102 may be modified to be embedded in the board.

The battery management device 102 may be configured to measure the difference in potential between the two ends of the battery 104, the surface concentration at the cathode and the anode, the solid potential at the anode, concentration, electrolyte potential, and the like of the battery 104. For example, The battery management device 102 may provide probabilistic information of an unobservable variable such as a state of charge. For example, the vehicle control system including the controller of the autonomous traveling / semi-autonomous traveling vehicle may be integrated with the battery management apparatus 102, or the control system of the object Internet apparatus may be integrated with the battery management apparatus 102, .

The battery management apparatus 102 can accurately track the internal state of the battery 104 within 160 seconds from the start of charge / discharge of the battery 104 when the initial charge state is not known. The battery management apparatus 102 can accurately predict the internal state of the battery 104 over the entire battery 104. [ For example, the battery management device 102 may handle any change in the load current at the load current associated with the actual drive cycle or handle changes in the operating voltage or operating current of the object Internet device.

A method for estimating a state of charge of a battery according to an embodiment of the present invention is based on a long-term energy (LTE) estimation algorithm, a short-time energy (STE) estimation algorithm, and a terminal voltage measurement, And a provisional correction step between the long-term energy and the short-time energy to improve the long-term energy. This can be done by applying the properties of the estimated parameter of charge state as a function of time. In other words, the amount of change in the charge state can be predicted based on the difference between the average terminal voltages and the similarity between the difference values of the open circuit voltages of the long-term energy estimation algorithm. Based on the prediction, the state of charge may be updated after inverse Coulomb counting.

In the short-term energy estimation algorithm, the discharge currents within a short period of time are estimated by using the characteristics at the point where the open-circuit voltage and the resistance value slowly change as a function of time in a short period of time, Can be predicted. Thereafter, the state of charge can be predicted using the coulomb counting process.

Finally, the provisional correction method can be proposed for the provisional correction of errors that can be scaled in a continuous process that fixes the previous value in the long-term applied short-term energy estimation method.

The charged state energy estimation method of the present invention includes a low frequency band filter in a preprocessing process for eliminating various power noise and reducing occurrence of estimation error, which can be generated in an environment for measuring intermittent terminal voltage .

2 is a flowchart illustrating a long-term energy estimation process according to an embodiment of the present invention.

The present invention can extend the estimated unit time to correct the estimation error of the state of charge caused by the large fluctuation of the terminal voltage in the fluctuating discharge state which frequently occurs in recent matter Internet devices. Further, it is possible to calculate the differences between the open-circuit voltage and the charged state based on the difference in the terminal voltage for estimating the average discharge current by applying the inverse coulomb counting process. Thereafter, the open-circuit voltage value can be predicted by calculating the difference between the currently measured terminal voltage Vt [n] and the voltage drop (IR).

First, it is determined whether a flag indicating whether an open-circuit voltage value OCV_RST exists in a relaxation state is 1 (S10). If there is an open-circuit voltage value OCV_RST, the open-circuit voltage value OCV _{RST in the} idle state is set to the open-circuit voltage value OCV [nd] in the nd time (S11). On the other hand, if the open-circuit voltage value OCV_RST does not exist, it is determined whether the initialization flag FLAG _INT indicating whether the initialization step is being performed is 1 (S12). If the initialization flag is 1, the resistance value R _LTE [n] at n is set to the resistance value R _LTE [n-1] at n-1 (S13). On the other hand, when the initialization flag is 0, the resistance value _RSTE [n] in n of the short-term energy estimation process is set to the resistance value _RSTE [n-1] at n-1 (S14) .

Then, the open-circuit voltage value OCV [n] at n can be calculated using the difference value of the terminal voltage (DELTA Vt = DELTA OCV) and the difference value of the terminal voltage (S15). It is possible to calculate the state of charge (SOC _LTE [nd], SOC _LTE [n]) of the long-term energy process in nd and n using the table SOC_TBL of the state of charge of the open- The difference value? SOC of the charge state value can be measured (S16). Further, the average current value I _avr can be calculated using the difference value of the charge state value, the maximum charge amount Q _max , and the measurement time (S17), and the average current value and the average current value to the terminal voltage value (Vd) by using the resistance values, the temporary open-circuit voltage (OCV _tmp) by using the terminal voltage value, the maximum open-circuit voltage (OCV _max) and the temporary open-circuit voltage (OCV _tmp) It is possible to predict the open-circuit voltage value OCV [n] at n (S18). Finally, the obtained open-circuit voltage value (OCV [n]) at n is substituted into the charge state value table to obtain the charge state value SOC _LTE [n] at n, The resistance value R _LTE [n] at n can be obtained by substituting it into the table (S19).

This method can obtain the initial value after the operation for a specific time without referring to the initial value, and can predict the average change of the state of charge in the long-term, so that the error occurrence problem accumulated in the usual coulomb counting method is corrected can do.

The change in signal related to battery status over time can be seen in the Charge State-Open-Circuit Voltage (SOC-OCV) and Charge State-Resistance (SOC-R) graphs. When analyzing these frequency characteristics, the open-circuit voltage and internal resistance parameters changed very slowly over time compared to the terminal voltage. This implies that the open-circuit voltage and resistance values can be approximated in the process of predicting the short-time discharge current. When the difference between the OCV [n-1] value and the Vt [n] value can be divided into R [n-1] values based on a change in this characteristic in the battery signal associated with the charge time measurement, the discharge current I [ n]) can be predicted within a small error range.

The following equations (1) and (2) represent a process of estimating the current open-circuit voltage and resistance using previous values, and the following equation (3) is a formula for calculating I [n] based on this assumption.

FIG. 3 is a flow chart illustrating a short-term energy (STE) estimation process in accordance with an embodiment of the present invention, which has been proposed based on the above-described assumptions.

Referring to Figure 3, it can be seen, first, if the at rest the flag indicating whether the open-circuit voltage value at the rest condition is present open-circuit voltage flag _{(FLAG OCV _ RST) 1 (} S20). If the open-circuit voltage flag is 1, the open-circuit voltage value OCV _{RST in} the idle state can be set to the open-circuit voltage value OCV [n-1] at n-1 (S22). On the other hand, if the open-circuit voltage flag is 0, the open-circuit voltage value OCV [n] at n can be set to the open-circuit voltage value OCV [n-1] at n-1 (S21).

Thereafter, the long-term flag FLAG _LTE indicating whether or not the long-term energy estimation process is being performed can be confirmed (S23). If said if a long term flag is 1, the charge state value of the long-term energy estimation process in n (SOC _LTE [n]) and the resistance (RLTE [n]) the state of charge value at n-1 (SOC _STE [ (SOC _STE [n]) and the resistance value _RSTE [n-1] in the short-term energy estimation process at n when the long-term flag is 0, value (R _STE [n]) to be set to a state of charge value _{(SOC STE [n-1]} ) and the resistance _{(R STE [n-1]} ) of the short-term energy estimation process in the n-1 (S24 ).

The current value I [n] at n and the clipped current value Ic [n] are calculated using the charge state value SOC [n-1] and the resistance value R [n- n]) (S26). The current value I [n] at n can be calculated using Equation (3). Further, the charge state value SOC _STE [n] at n can be obtained using the clipped current value Ic [n] (S27). The long-term flag FLAG _LTE is checked again (S28). If the long-term flag FLAG _LTE is 1, the charge state value SOC _STE [n] in the n is stored in the charge state value table SOC_TBL applied can obtain a resistance value obtained (R _STE [n]), and the resistance value (R _STE [n]) open-circuit voltage (OCV [n]) at n, using the at n (S29 ). Further, by using the open-circuit voltage (OCV [n]) at the n-obtained charge value (SOC _STE [n]) at n again, using the state of charge value (SOC _STE [n]) a resistance value (R _STE [n]) at n can be obtained again (S29_5).

The calculated I [n] value can be classified through the clipping process by using the fact that the open-circuit voltage value is always larger than the terminal voltage value, and therefore only the positive value remains. The Ic [n] value already calculated through clipping can then be processed by the inverse coulomb counting method to calculate the shortturn change in the charged state. The current charge state value may then be estimated by subtracting the changed charge state value from the previous charge state value. Finally, the current R and open-circuit voltage values can be updated using the SOC-R and SOC-OCV tables, respectively.

This method can increase the accuracy to estimate the state of charge in the short-term. However, if the discharge state of the variable battery continues without a dormant state, there may be a problem that errors accumulated and accumulated in the process of estimating the open-circuit voltage and resistance value, which are the previous values, are constantly increased.

The estimation method according to an embodiment of the present invention combines the long-term energy and the short-term energy estimation processes in order to perform a provisional correction on the accumulated error problem in the short-term energy estimation process. The average difference between the terminal voltage values is similar to the difference between the open-circuit voltage values, so that the accumulated charge state estimation error due to the subsequent short-term energy estimation process can be estimated through the long-term energy and the short- . ≪ / RTI > Since the current discharged from the shorttom is more accurate to be predicted by the shorttom energy estimation process, the longtomem and shorttom energy estimation processes can be performed continuously in a provisional correction process.

The charge state value (SOC _LTE ) and the resistance value (R _LTE ) obtained in the long-term energy estimation process are transmitted to the short-term energy estimation process, and only a short-term energy estimation process can be performed thereafter. 3, the charge state value (SOC _LTE ) and the resistance value (R _LTE ) value transmitted from the long-term energy estimation process, when the provisional correction mode of the long-time energy and the short-term energy estimation process are large, And the instantaneous discharge current can be calculated using the previous value, which is shown in Equation (3). Thereafter, the short-term difference value in the charge state can be obtained using the coulomb counting method, and when the flag of the current operation mode is in the provisional correction state of the long-term energy and the short-term energy estimation process, R _STE ) value can be updated and a voltage drop can be calculated. The final open-circuit voltage value can be obtained by adding IR to the current terminal voltage. In the final stage, all SOC_STE and R_STE values can be updated based on the obtained open-circuit voltage value.

4 is a flowchart illustrating a method of estimating a state of charge of a battery according to an embodiment of the present invention.

Referring to FIG. 4, a method for estimating a state of charge of a battery according to an embodiment of the present invention includes an initialization step, a long-term energy estimation step, a short-term energy estimation step, and a dormant step.

First, the estimation method determines whether the battery mounted on the battery-powered device is used for the first time without being used previously (S30). If the battery is used for the first time, an initialization step may be initiated to set an initial charge state value, an initial terminal voltage value, and an initial resistance value (S31). When the initialization step S31 is completed, the long-term energy estimation process can be started (S32). Through the long-term energy estimation process, a long-term flag FLAG _LTE indicating a charge state value (SOC _LTE ), a resistance value (R _LTE ), and a long-term energy estimation process in the long-term energy estimation can be obtained S33).

If it is determined in step S30 that the battery is used for the first time, the short-term energy estimation process may be started in step S34. If the battery has been used, the current charge state value, the terminal voltage value, and the resistance value, which are all information on the battery, are all known, so that the process can be performed based on them. It is possible to determine whether the current value predicted during the short-term energy estimation process is smaller than the minimum current value and the terminal voltage value is a stable value for a predetermined time or longer (S35).

If the condition is not satisfied, the short-term energy estimating process determines whether the charge state value (SOC _STE ), resistance value (R _STE ), open-circuit voltage value (OCV _STE ) (FLAG _STE ) indicating whether or not the flag is set (S36). Further, if the above condition is satisfied, the battery can be introduced into a rest state (S37). If it is determined that the current value is smaller than the minimum current value and the terminal voltage value is stable for more than a predetermined period of time (S38) by predicting the current value even during the idle state, the idle state (S37) . If the above condition is not satisfied, the open-circuit voltage value (OCV) can be obtained (S39).

5 is a flowchart showing a detailed method for estimating a state of charge of a battery according to an embodiment of the present invention.

Referring to FIG. 5, the sequence includes an initialization step, an LTE estimation step, an STE estimation step, and a dormant step. For the system in which the proposed method is embedded, the initialization step may be performed differently depending on whether the currently connected battery is turned on, whether the battery is used, or whether it is new (S50). If a new battery is connected, the existing information about the battery is not known at all, so an initialization step for determining the initial value should be started (S51). If the currently connected battery is a previously used battery, the system already knows all the information about the state of the battery (SOC, OCV, R, etc.), so the limiting algorithm process can estimate the state of charge for a short time And may be moved to the STE (S59, S60).

In the initialization step, the initial terminal voltage may be updated using the OCV-SOC table, and the LTE process may be started (S52). In the LTE process, the initial charge state value can be estimated by repeating the process until the SOC_LTE difference value is greater than the SOC_TH0 value.

The initial value SOC, R is estimated in this way and the associated FLAG value FLAG_LTE can be transmitted to the short-term energy estimation process. 2, if the difference value of the charge state value is greater than the threshold value (Yes in S55), a provisional correction process is performed (S58), and the charge state Value, a resistance value, and a flag value, which can be transmitted to the short-term energy estimation process.

The process for estimating the short-term energy (S60) can also be performed with reference to FIG. It is checked whether the predicted current value is smaller than the minimum current value and the voltage is stable during the short-term energy estimation process (S61). If it is not the condition to check, it can be checked whether the measurement time exceeds 400 seconds ). If the measurement time exceeds 400 seconds, the short-term energy estimation process can be completed and the resistance value (R _STE ) can be obtained.

If the above conditions are satisfied in the step S61 of checking the current value and the voltage value, the current value and the voltage value may be introduced into the dormant step (S66). The current value and the voltage value are checked during the dormant step (S67). If the condition is satisfied, the open-circuit voltage value and the flag in the dormant step are compared with the short-term energy estimate To the start of the process (S68). While the dormant state is maintained, the terminal voltage and the open-circuit voltage value become substantially equal to each other, and thus the open-circuit voltage RST value is transmitted to the LTE and STE processes to continuously reset the open- . In the idle state, the discharge current I can be estimated using Equation (3). The I_min value determined in the idle state can be determined experimentally.

The charge state estimation algorithm according to an embodiment of the present invention can be embedded in a desktop computer using the C language. The experimental environment for battery discharge simulations may include a Keithley 2281s battery simulator, a Keithley SM2602 source meter for input of discharge current, and a Keithley DMM7510 multimeter for measurement of terminal voltage. In these experiments, a highly accurate simulation environment can be provided that is similar to what is possible in actual battery discharge conditions.

The initial modeling process in the 2281s can be performed using a BL-54SG 2610 mAh lithium-ion battery for the LG G2 smartphone. In order to create various discharge current profile scenarios in the environment of the latest small mobile devices and object internet devices, the discharge current input waveform signal was generated by MATLAB and 2281s was applied at a sampling rate of 20Hz using SM2602A. In addition, to obtain terminal voltage values that are measured intermittently in an actual variable discharge state, the DMM 7510 can be connected to 2281s to measure terminal voltage values at a 1 Hz sampling rate. These can be used as an input to the charge state estimation algorithm in a desktop environment.

Two experiments were conducted to confirm the accuracy of the estimation of the state of charge. The first evaluated the error range in the initial charge state estimation based on the LTE method according to one embodiment of the present invention in stable and variable discharge scenarios. Second, we compared the charge state estimation accuracy of the method according to one embodiment of the present invention and the general algorithm with the absolute value of the charge state error in a continuous stable / variable discharge scenario without dormancy.

In the first experimental example, the accuracy of the initial charge state estimation according to the LTE algorithm method of the present invention was confirmed. According to the experimental conditions for the initial value, the initial charge state of the new battery was estimated to be about 95%. The discharge current scenario for the experiment can be classified into a stable discharge state and a variable discharge state. The SOC_TH0 value for determining the timing of the initial value estimation in the LTE algorithm may be determined experimentally according to the type of the discharge scenario.

6A and 6B are graphs of experimental results for estimating initial charge state values in a stable discharge state.

FIG. 6A shows the input state of a stable staircase wave, and FIG. 6B shows the result of the LTE estimation method operating continuously for a specific time gradually reaching the state of charge of the simulator.

When the absolute value of SOC_LTE of the LTE algorithm becomes larger than TH1 (7%) at 649 seconds, an initial value can be estimated. At this moment, the state of charge of the simulator is about 89.2%, the SOC estimated by the LTE method may reach about 93.9%, and the accuracy level of the initial value estimation is within the error range of 4.7%.

FIGS. 7A and 7B are graphs of experimental results obtained by estimating initial charge state values in a variable discharge state.

FIG. 7A shows the input state of the variable discharge current (3.25 Hz wave), and FIG. 7B shows the LTE initial value estimation result. In order to estimate the initial value in the battery state during the fluctuating discharge, the LTE method was operated repeatedly for a longer time than in the estimation of the battery state value in the stable state. At 1144 seconds, when the SOC_LTE value was greater than TH2 (15%), the initial value was estimated. At that moment, the state of charge of the simulator was 79.9%, and the state of charge estimated by the LTE method reached 77.3%. Also, the accuracy level for estimating the initial value was about 2.6% of the error range.

As a result, the general initial value estimation method can estimate the initial value only when it is introduced into the idle state. On the other hand, experimental results on initial value estimation indicate that the method according to an embodiment of the present invention can roughly estimate the state of charge by performing the estimation process for a specific time.

In the second experimental example, the charge state was estimated on a continuous discharge scenario without dormancy for a long time, which is a frequent occurrence in today's small mobile and things Internet devices. To measure the estimation accuracy in the various discharge states, both the stable discharge state and the variable discharge state scenario were applied to the experiment. In addition, the accuracy of the charge state estimation by the general methods and the proposed method can be compared to the absolute value of the error for the charge state.

8A and 8D show experimental results in a scenario of a stable discharge current for 2036 seconds.

Fig. 8A shows the input scenario of the stepped waveform discharge current, Fig. 8B shows the measured terminal voltage value at a 1 Hz sampling rate filtered by a 50-tap moving average filter (MV50), Fig. As shown in FIG. Also, FIG. 8D shows the absolute values of the errors accumulated in the estimation of the state of charge of each algorithm.

8D and the results of the experimental example shown in the following Table 1, since this is measured before the performance of the provisional correction of the LTE, if the charge state estimation error of the proposed method is equal to the estimation error of the STE method, And 0.035%, respectively. Since the provisional correction process is applied through the LTE process at 1167 seconds, the estimation of the charge state of the proposed method may be 0.052 to 0.068% less than the STE method. Estimation errors of the FTV method can be 0.042% less than that of the STE method in 2036 seconds, but the estimation error of the proposed method can be 0.001% less than the FTV method. In conclusion, the estimation accuracy of the proposed method can be the highest throughout the discharge time.

Time
(sec) 2281s_
SOC FTV_
eSOC | FTV_eSOC_
Err | STE_
eSOC | STE_eSOC_
Err | The proposed eSOC | Proposed eSOC_Err | 700 93.754 93.636 0.118 93.670 0.083 93.670 0.083 1167 89.592 89.517 0.074 89.534 0.058 89.586 0.006 2036 81.858 81.841 0.017 81.799 0.059 81.875 0.016

Figures 9A-9D and Table 2 below show charge state estimation errors for a variable input waveform (3.25 Hz). Generally, both the FTV and STE methods are continuous because the estimation process is continuous, resulting in a continuous increase in the error that is generated and accumulated. On the other hand, the charging state estimation error of the proposed method can be reduced because of the provisional correction process. At 2530 second, the estimation error of the proposed method can be 0.283% smaller than the STE method and 0.171% smaller than the FTV method.

In the case of 3689 seconds in which the discharge is sustained for a long time, the estimation error of the proposed method may be 0.539% lower than that of the STE method and 0.362% lower than that of the FTV method.

Time
(sec) 2281s_
SOC FTV_
eSOC | FTV_eSOC_
Err | STE_
eSOC | STE_eSOC_
Err | The proposed eSOC | Proposed eSOC_Err | 1232 83.799 83.053 0.746 83.034 0.765 83.113 0.686 2530 66.761 65.612 1.149 65.499 1.261 65.783 0.978 3698 51.424 49.857 1.567 49.680 1.744 50.219 1.205

FIGS. 10A to 10D and Table 3 below represent charge state estimation errors in a 6.25 Hz input waveform scenario, which is performed at high frequencies. In general, the accumulated error is increased due to a large margin as compared with the previous experiment, and the charge state estimation error of the proposed method can be minimized. At 4800 second, the estimation error of proposed method is 0.968% less than STE method and 0.798% less than FTV method.

Time
(sec) 2281s_
SOC FTV_
eSOC | FTV_eSOC_
Err | STE_
eSOC | STE_eSOC_
Err | The proposed eSOC | Proposed eSOC_Err | 1600 77.385 76.508 0.877 76.470 0.915 76.597 0.788 3200 54.758 53.349 1.409 53.229 1.530 53.731 1.027 4800 32.141 30.087 2.054 29.916 2.224 30.885 1.256

The method of estimating the state of charge of a battery according to an embodiment of the present invention is an increase in the general method of estimating the state of charge using only the terminal voltage in the continuous variable discharge state frequently occurring in small portable and object Internet devices Can improve the accumulated error problem. For this purpose, we propose a potential correction method that analyzes both the deformation characteristics of parameters related to the battery state in the time function and uses both STE and LTE estimation algorithms in combination. Also, the method of estimating the state of charge of a battery according to an embodiment of the present invention includes upgrading a general estimation method in which an approximate initial state of charge value only waits for a dormant state to determine initial values, so that accurate estimation through an LTE- . The estimation error of the state of charge of the proposed method may be less than the other methods when the accuracy of the method can be estimated by applying it to long term continuous stable discharge and variable discharge scenarios.

However, the method of estimating the state of charge of a battery according to an embodiment of the present invention has a limitation that it is difficult to accurately and accurately estimate an initial value, and is not appropriately considered in the case of a multi-cell. Methods for estimating initial values with higher accuracy in cell balancing as well as charging states of multi-cell batteries will be studied in the future.

It will be apparent to those skilled in the art that various modifications and variations can be made in the present invention without departing from the spirit or scope of the invention as defined in the appended claims. It will be clear to those who have knowledge.

Claims (13)

Determining whether the battery mounted on the apparatus using the battery is used for the first time;
Performing the initialization if the battery is first used;
Performing the long-term energy process when performing the initialization;
Estimating a charged state value and a resistance value of the battery through the long-term energy process;
Performing the short-time energy process if the initialization is not performed;
Determining whether to introduce the battery into a dormant state based on a current value and a voltage value of the battery;
The method comprising: acquiring an open-circuit voltage value of the battery to re-predict a state of charge of the battery when the battery is introduced into the dormant state; And
Further comprising the step of estimating a current resistance value and a current open-circuit voltage value of the battery by continuously performing the short-time energy process when the battery is not introduced into the dormant state. The method according to claim 1,
Wherein performing the initialization comprises setting an initial terminal voltage, an initial charge state value, and an initial resistance value. The method according to claim 1,
Wherein performing the long-term energy process comprises:
Estimating an open-circuit voltage value using the initial terminal voltage value and the measured terminal voltage value;
Measuring a charge state difference value using the open-circuit voltage value, the initial charge state value, and the measured charge state value;
Estimating an average current value using the charge state difference value;
Predicting the open-circuit voltage value using a resistance value predicted based on the average current value and the initial resistance value; And
Estimating a charge state value and a resistance value of the long-term energy based on the re-predicted open-circuit voltage value and the predicted resistance value. The method according to claim 1,
Wherein performing the short-term energy process comprises:
Estimating an open-circuit voltage, a charge state value, and a resistance value of the battery;
Estimating a current value using the open-circuit voltage, the charge state value, and the resistance value; And
And estimating a current charging state value and a current resistance value of the battery using the current value and the difference value of the charging state value. The method according to claim 1,
Wherein the step of determining whether to introduce the battery into the dormant state includes the steps of determining whether the current value of the battery is less than the minimum current value or the current terminal voltage value is maintained at a constant value longer than the threshold time, / RTI > The method according to claim 1,
Wherein the step of re-predicting the state of charge of the battery further comprises performing at least one of the long-time energy process and the short-time energy process again. A computer-readable recording medium recording a program for executing the method of any one of claims 1 to 6 in a computer. A battery management apparatus comprising:
Wherein the battery management apparatus includes a processor,
Determining whether the battery mounted on the apparatus using the battery is used for the first time;
Performing an initialization if the battery is first used;
Performing the longtomes energy process when performing the initialization;
Predicting a charged state value and a resistance value of the battery through the long-term energy process;
Performing the short-time energy process when the initialization is not performed;
Determining whether to introduce into a dormant state based on a current value and a voltage value of the battery;
Acquiring an open-circuit voltage value of the battery to re-predict the state of charge of the battery when the battery is introduced into the dormant state; And
Wherein the controller is configured to continuously perform the short-time energy process to predict a current resistance value and a current open-circuit voltage value of the battery when the battery is not in the idle state. 9. The method of claim 8,
Wherein the initializing operation sets an initial terminal voltage, an initial charge state value, and an initial resistance value. 9. The method of claim 8,
Wherein the act of performing the long-term energy process comprises:
Obtaining an open-circuit voltage value by using the initial terminal voltage value and the measured terminal voltage value to estimate a state of charge of the battery;
Measuring a charge state difference value using the open-circuit voltage value, the initial charge state value, and the estimated charge state value;
Predicting an average current value using the charge state difference value;
Obtaining an open-circuit voltage value using a resistance value predicted based on the average current value and the initial resistance value, and re-predicting a state of charge of the battery; And
And estimating a charge state value and a resistance value of the long-term energy based on the predicted open-circuit voltage value and the predicted resistance value. 9. The method of claim 8,
The act of performing the short-
Estimating an open-circuit voltage, a charge state value, and a resistance value of the battery;
Estimating a current value using the open-circuit voltage, the charge state value, and the resistance value; And
And estimating a current charging state value and a current resistance value of the battery using the current value and the difference value of the charging state value. 9. The method of claim 8,
Wherein the step of determining whether to introduce the battery into the dormant state introduces the battery into the dormant state when the current value of the battery is less than the minimum current value or the period in which the current terminal voltage value is kept constant is longer than the threshold time. 9. The method of claim 8,
Wherein the step of re-predicting the state of charge of the battery comprises performing at least one of the long-time energy process and the short-time energy process again.

Images