Method and apparatus for detecting charged state of secondary battery based on neural network calculation

• US 20060181245A1
• Filed: 02/14/2006
• Published: 08/17/2006
• Est. Priority Date: 02/14/2005
• Status: Active Grant

• Current Assignee: DENSO Corporation, Nagoya Institute of Technology, Nippon Soken, Inc.
• Original Assignee: DENSO Corporation, Nagoya Institute of Technology, Nippon Soken, Inc.
• Inventors: Morita, Yoshifumi, Kozawa, Takaharu, Ichikawa, Atsushi, Mizuno, Satoru, Sakai, Shoji, Hashikawa, Atsushi, Mizuno, Naoki
• Granted Patent: US 7,554,296 B2
• Time in Patent Office: Days
• US Class Current: 320/132
• CPC Class Codes: G01R 31/3648 comprising digital calculation means, e.g. for performing an algorithm
• CPC Class Codes: G01R 31/367 Software therefor, e.g. for battery testing using modelling or look-up tables
• CPC Class Codes: G01R 31/3842 combining voltage and current measurements

Abstract

A neural network type of apparatus is provided to detect an internal state of a secondary battery implemented in a battery system. The apparatus comprises a detecting unit, producing unit and estimating unit. The detecting unit detects electric signals indicating an operating state of the battery. The producing unit produces, using the electric signals, an input parameter required for estimating the internal state of the battery. The input parameter reflects calibration of a present charged state of the battery which is attributable to at least one of a present degraded state of the battery and a difference in types of the battery. The estimating unit estimates an output parameter indicating the charged state of the battery by applying the input parameter to neural network calculation.

Claims

1. A neural network type of apparatus for detecting an internal state of a secondary battery implemented in a battery system, the apparatus comprising:

detecting means for detecting electric signals indicating an operating state of the battery;

producing means for producing, using the electric signals, an input parameter required for estimating the internal state of the battery, the input parameter reflecting calibration of a present charged state of the battery; and

estimating means for estimating an output parameter indicating the charged state of the battery by applying the input parameter to neural network calculation.

2. The apparatus of claim 1, wherein

the electric signals are voltage and current of the battery acquired in real time during a predetermined period of time;

the input parameter consisting of a first input parameter indicating the operating state of the battery and a second input parameter indicating the degraded state of the battery; and

the producing means comprises

means for calculating the first input parameter on the basis of data of the voltage and current of the battery, and

means for calculating the second input parameter in response to a state of predetermined charge of the battery.

3. The apparatus of claim 2, wherein

the outputting parameter estimating means is means for calculating the output parameter indicating a present charged state of the battery by processing both the first and second input parameters based on a neural network calculation technique.

4. The apparatus of claim 3, wherein

the first input parameter includes data indicative of history of the voltage detected during the predetermined period of time.

5. The apparatus of claim 4, wherein

the first input parameter includes data of history of both the voltage and the current detected during the predetermined period of time.

6. The apparatus of claim 5, wherein

the second input parameter calculating means includes

means for calculating the data of the voltage history and the current history into an approximate expression on a least-squares method;

and

means for calculating at least one of a present value of an open-circuit voltage of the battery and a present value of an internal resistance of the battery on the basis of the approximate expression, both the present values being included in the first input parameters, and

the output parameter calculating means is configured to calculate the output parameter by using both the first parameter consisting of at least one of the data of the voltage history, the data of the current history, the present value of the open-circuit voltage and the present value of the internal resistance and the second input parameter.

7. The apparatus of claim 3, wherein

the second input parameter indicating the degraded state of the battery is an open-circuit voltage of the battery detected in response to a state of predetermined charge of the battery and

the output parameter indicating the charged state of the battery is an SOC (state of charge) of the battery.

8. The apparatus of claim 3, wherein

the second input parameter indicating the degraded state of the battery is a difference between an open-circuit voltage of the battery detected when the battery is in a substantially full charge state and the open-circuit voltage of the battery detected in response to a state of predetermined charge of the battery and

the output parameter indicating the charged state of the battery is an SOC (state of charge) of the battery.

9. The apparatus of claim 3, wherein

the second input parameter indicating the degraded state of the battery is composed of (i) a difference between an open-circuit voltage of the battery detected when the battery is in a substantially full charge state thereof and the open-circuit voltage of the battery detected in response to a state of predetermined charge of the battery and (ii) the open-circuit voltage of the battery detected in response to a state of predetermined charge of the battery and

the output parameter indicating the charged state of the battery is an SOC (state of charge) of the battery.

10. The apparatus of claim 3, wherein

the second input parameter indicating the degraded state of the battery is a value of internal resistance of the battery detected in response to a state of predetermined charge of the battery and

the output parameter indicating the charged state of the battery is an SOH (state of health) of the battery.

11. The apparatus of claim 3, wherein

the second input parameter calculating means includes

means for calculating data of history of both the voltage and the current into an approximate expression on a least-squares method;

and

means for calculating a present value of an open-circuit voltage of the battery on the approximate expression, the present value being included in the first input parameters, and

the output parameter calculating means is configured to calculate the output parameter by using both the first parameter consisting of the data of the voltage history, the data of the current history, and the present value of the open-circuit voltage and the second input parameter.

12. The apparatus of claim 11, wherein

the second input parameter indicating the degraded state of the battery is both an open-circuit voltage and an internal resistance of the battery detected in response to a state of predetermined charge of the battery.

13. The apparatus of claim 12, wherein

the output parameter indicating the present charged state is one of an SOC (state of charge) of the battery, an SOH (state of health) of the battery, and a function whose variables including information indicative of the SOC and SOH.

14. The apparatus of claim 13, wherein

the function is a degree of degradation of the battery which is defined by an expression of

“

the degree of desegregation=SOH/(initial full charge capacity×

SOC).”

15. The apparatus of claim 13, wherein

the output parameter calculating means is configured to allow the SOC, the open-circuit voltage, and the internal resistance to be subjected to neutral network calculation so as to calculate the SOH.

16. The apparatus of claim 2, wherein

the first input parameter calculating means is configured to calculate, as the first input parameter, voltage history data and current history data based on data of the received voltage and current of the battery;

the second input parameter calculating means is configured to calculate, as the second input parameter, an open-circuit voltage of the battery and an internal resistance of the battery using both the voltage history data and the current history data in response to a state of predetermined charge of the battery; and

the outputting parameter estimating means is means for calculating the output parameter indicating a full charge capacity of the battery, the full charge capacity being expected at present.

17. The apparatus of claim 16, wherein

the second input parameter calculating means is configured to use a least-squares method so as to create an approximate linear expression from both the voltage history data and the current history data and to calculate both the open-circuit voltage and the internal resistance using the approximate linear expression.

18. The apparatus of claim 17, further comprising

means for calculating a degree of degradation of the battery based on an expression of:

DD=Q_present/Q_initial

, wherein DD denotes the degree of degradation of the battery, Q_present denotes a present value of the full charge capacity estimated by the output parameter calculating means, and Q_initial denotes an initial value of the full charge capacity given when the battery is manufactured.

19. The apparatus of claim 1, wherein

the electric signals are voltage and current of the battery acquired in real time during a predetermined period of time;

the estimating means comprises a neural network having the input layer which receives the input parameter and an intermediate layer which performs neural network calculation using both of the input parameters and rewritable coupling coefficients mutually connecting the input layer, the intermediate layer, and an output layer so as to estimate the output parameter and allow the output parameter to be outputted from the output layer, the output parameter indicating the charged state of the battery; and

the producing means comprises

a memory with a plurality of memory tables each storing the coupling coefficients, the coupling coefficient stored in the memory tables being different from one another table by table and depending on divided ranges of a characteristic of a specified input parameter specified in the input parameter, the specified input parameter having a correlation with a degraded state of the battery; and

selecting means for selecting, in response to a present value of the specified input parameter, a memory table from the plurality of memory tables to read in the coupling coefficients stored therein, the specified coupling coefficients being given to the neural network calculation.

20. The apparatus of claim 19, wherein

the input parameters include voltage history data, current history data, and an open-circuit voltage of the battery which are estimated on the voltage and current of the battery and

the specified input parameter is an open-circuit voltage of the battery to be obtained when the battery is in a fully charged state thereof.

21. The apparatus of claim 19, wherein

the input parameters include voltage history data, current history data, and an open-circuit voltage of the battery which are estimated on the voltage and current of the battery and

the specified input parameter is an open-circuit voltage of the battery to be obtained in response to a state of predetermined charge of the battery.

22. The apparatus of claim 20, wherein

the open-circuit voltage of the battery in response to the state of predetermined charge of the battery has a voltage characteristic which is divided into plural voltage ranges depending on a magnitude of the voltage and

the coupling coefficient stored in the memory tables are different from one another table by table and depending on a magnitude of the open-circuit voltage of the battery in response to the state of predetermined charge of the battery

23. The apparatus of claim 1, wherein

the producing means is configured to produce the input parameter which is calibrated depending on a charge and discharge characteristic of the battery which is attributable to at least one of the degraded state of the battery and the difference in types of the battery.

24. The apparatus of claim 23, wherein

the input parameter includes either a voltage V of the battery or a ratio V/Vf wherein Vf is a voltage of the battery detected when the battery is in a fully charged state, either an open-circuit voltage Vo of the battery or a ratio of Vo/Vof wherein Vof is an open-circuit voltage detected when the battery is in a fully charged state, either an internal resistance R of the battery or a ratio of R/Rf wherein Rf is an internal resistance detected when the battery is in a fully charged state, a predetermined function f(Vo, R) using, as input variables, the open-circuit voltage Vo and the internal resistance R and expressing a correlation to an amount of current of the battery which is dischargeable at present, and a current I of the battery.

25. The apparatus of claim 24, wherein

the function f(Vo, R) is a function of which input variable is based at least on a ratio of Vo/R.

26. The apparatus of claim 25, wherein

the function f(Vo, R) is a function of which input variable is based on a ratio of Vo·

Vo/R.

27. The apparatus of claim 25, wherein

the function f(Vo, R) is a function of which input variable is based on a ratio of (Vm−

Vo)/R wherein Vm is a predetermined voltage of the battery.

28. The apparatus of claim 25, wherein

the predetermined voltage Vm is a discharge stop voltage of the battery.

29. The apparatus of claim 24, wherein

the function f(Vo, R) is a function defined by f(Vop, Rp)/f(Vof, Rf) wherein f(Vop, Rp) denotes a present value of the function f(Vo, R) and f(Vof, Rf) denotes a value of the function f(Vop, Rp) obtained when the battery is in a fully charged state.

30. The apparatus of claim 29, wherein

the function f(Vo, R) is a function defined by (Vo·

Vo/R)/(Vof·

Vof/Rf) wherein Vof and Rf denote an open-circuit voltage and an internal resistance of the battery detected when the battery is in a fully charged state, respectively.

31. The apparatus of claim 1, wherein

the input parameter is composed of a plurality of physical state quantities reflecting the present charged state of the battery and

the plurality of physical state quantities include a ratio between a present value of a physical state quantity specified among the physical state quantifies and a value of the specified physical state quantity acquired in response to a state of predetermined charge of the battery.

32. The apparatus of claim 31, wherein

the plurality of physical state quantities include at least one an average voltage Va, an open-circuit voltage Vo, an internal resistance R, and a maximum discharge power P of the battery.

33. The apparatus of claim 31, wherein

the plurality of physical state quantities include all of an average voltage Va, an open-circuit voltage Vo, an internal resistance R, and a maximum discharge power P of the battery.

34. The apparatus of claim 1, wherein

the input parameter is composed of a plurality of physical state quantities reflecting the present charged state of the battery and

the plurality of physical state quantities include a specified physical state quantity given by a linear function of which variables include a present value of the specified physical state quantity and a value of the specified physical state quantity acquired in response to a state of predetermined charge of the battery.

35. The apparatus of claim 34, wherein

the plurality of physical state quantities include at least one an average voltage Va, an open-circuit voltage Vo, an internal resistance R, and a maximum discharge power P of the battery.

36. The apparatus of claim 34, wherein

the plurality of physical state quantities include all of an average voltage Va, an open-circuit voltage Vo, an internal resistance R, and a maximum discharge power P of the battery.

37. The apparatus of claim 1, wherein the battery system is an on-vehicle battery system mounted on a vehicle.

38. A method of detecting an internal state of a secondary battery implemented in a battery system, comprising steps of:

detecting electric signals indicating an operating state of the battery;

producing, using the electric signals, an input parameter required for estimating the internal state of the battery, the input parameter reflecting calibration of a present charged state of the battery which is attributable to a present degraded state of the battery; and

estimating an output parameter indicating the charged state of the battery by applying the input parameter to neural network calculation.