METHOD AND APPARATUS ESTIMATING STATE OF BATTERY

• US 20160209473A1
• Filed: 09/24/2015
• Published: 07/21/2016
• Est. Priority Date: 01/21/2015
• Status: Abandoned Application

• Current Assignee: Samsung Electronics Co. Ltd. (Samsung Group)
• Original Assignee: Samsung Electronics Co. Ltd. (Samsung Group)
• Inventors: YOU, Kaeweon, PARK, SangDo
• Application Number:
  US14/863,792
• Publication Number:
  US 20160209473A1
• Time in Patent Office: Days
• US Class Current: 1/1
• CPC Class Codes: G01R 31/367 Software therefor, e.g. for battery testing using modelling or look-up tables
• CPC Class Codes: G01R 31/382 Arrangements for monitoring battery or accumulator variables, e.g. SoC
• CPC Class Codes: G01R 31/392 Determining battery ageing or deterioration, e.g. state of health

Abstract

A method and apparatus for estimating a state of a battery are provided. A battery life estimation apparatus may acquire sensing data of a battery, may extract a stress pattern from the sensing data that represents changes in states of the battery based on stress applied to the battery, and may estimate a life of the battery based on the stress pattern.

Claims

1. A battery life estimation apparatus comprising:

a stress pattern extractor configured to use at least one processing device to extract a stress pattern from sensing data acquired for a battery, the stress pattern representing changes in states of the battery based on stresses applied to the battery and characterized by categorizing different stresses represented in the sensing data; and

a life estimator configured to use at least one processing device to estimate a life of the battery based on the characterized stress pattern.

2. The battery life estimation apparatus of claim 1, further comprising a sensor system including a plurality of sensors to measure the sensing data of the battery, the sensing data being real time measurements of physical properties of the battery.

3. The battery life estimation apparatus of claim 1, wherein the life estimator estimates the life of the battery in real time by providing characteristic data, as the categorized different stresses, to a learner to which a learning parameter is applied, wherein the learning parameter is previously trained on battery training sensing data of a previous time.

4. The battery life estimation apparatus of claim 1, wherein the sensing data comprises at least one of voltage data, current data, and temperature data of the battery sensed from respective sensors configured to measure corresponding properties of the battery.

5. The battery life estimation apparatus of claim 1, wherein the stress pattern extractor is configured to extract the stress pattern from the sensing data using a rainflow counting scheme, and

wherein the stress pattern represents a plurality of cycles that respectively represent changes in values of the sensing data over time.

6. The battery life estimation apparatus of claim 5, wherein the stress pattern extractor is configured to perform the categorizing by extracting a level for each of the plurality of cycles from a plurality of levels of a determined parameter, and configured to generate, based on each of the levels, characteristic data representing a characteristic of the stress pattern.

7. The battery life estimation apparatus of claim 6, wherein the stress pattern extractor is configured to perform the categorizing by generating the characteristic data based on a determined number of cycles, of the plurality of cycles, that correspond to each of the plurality of levels.

8. The battery life estimation apparatus of claim 6, wherein the determined parameter comprises at least one of an offset, an amplitude, and a period of each of the plurality of cycles.

9. The battery life estimation apparatus of claim 6, wherein the stress pattern extractor is configured to create a plurality of combination parameters, each representing respective levels for each of a plurality of the determined parameters for a cycle, and configured to perform the categorizing by generating the characteristic data based on a determined number of cycles, of the plurality of cycles, whose determined parameters correspond to each of the plurality of combination parameters.

10. The battery life estimation apparatus of claim 9, wherein the stress pattern extractor is configured to determine the number of cycles by applying different weights to different cycle patterns of the plurality of cycles.

11. The battery life estimation apparatus of claim 10, wherein the different cycle patterns include full and half cycle patterns.

12. The battery life estimation apparatus of claim 9, further comprising a dimension transformer configured to reduce a dimension of the characteristic data,

wherein the life estimator is configured to estimate the life of the battery by inputting the characteristic data with the reduced dimension to a predetermined learner to which a predetermined learning parameter is applied.

13. The battery life estimation apparatus of claim 6, wherein the stress pattern extractor is configured to generate the characteristic data at a predetermined period, so that characteristic data is generated for plural predetermined periods.

14. The battery life estimation apparatus of claim 6, wherein the life estimator is configured to estimate the life of the battery by inputting the characteristic data to a predetermined learner to which a predetermined learning parameter is applied.

15. The battery life estimation apparatus of claim 14, further comprising a dimension transformer configured to reduce a dimension of the characteristic data,

wherein the life estimator is configured to estimate the life of the battery by inputting the characteristic data with the reduced dimension to the predetermined learner.

16. The battery life estimation apparatus of claim 14, further comprising a communication interface, wherein the life estimator is configured to receive the predetermined learning parameter from an external apparatus using the communication interface, and configured to apply the received learning parameter to the predetermined learner.

17. The battery life estimation apparatus of claim 14, further comprising a storage configured to store in advance the predetermined learning parameter,

wherein the life estimator is configured to obtain the predetermined learning parameter from the storage and apply the obtained predetermined learning parameter to the predetermined learner.

18. The battery life estimation apparatus of claim 1, wherein the life estimator estimates the life of the battery in real time by providing characteristic data, as the categorized different stresses, to a learner to which a learning parameter is applied, and wherein the learning parameter is trained on battery training sensing data of a previous time, the life estimation apparatus further comprising:

a training data acquirer configured to acquire the battery training sensing data for the battery, in the previous time;

a training stress pattern extractor configured to use at least one processing device to extract a training stress pattern from the battery training sensing data, the training stress pattern representing changes in states of the battery based on stresses applied to the battery and characterized by categorizing different stresses represented in the training data; and

a learning parameter determiner configured to use at least one processing device to determine the learning parameter based on the characterized training stress pattern.

19. A battery life estimation apparatus comprising:

a training stress pattern extractor configured to use at least one processing device to extract a training stress pattern from training data for a battery, the training stress pattern representing change in states of the battery based on stresses applied to the battery and characterized by categorizing different stresses represented in the training data; and

a learning parameter determiner configured to use at least one processing device to determine a learning parameter based on the characterized training stress pattern, the learning parameter being determined for use in estimating a life of the battery.

20. The battery life estimation apparatus of claim 19, wherein the training data is derived from a previous measuring of physical properties of the battery.

21. The battery life estimation apparatus of claim 19, wherein the training stress pattern extractor is configured to extract the training stress pattern from the training data using a rainflow counting scheme, and

wherein the training stress pattern represents a plurality of cycles that respectively represent changes in values of the training data over time.

22. The battery life estimation apparatus of claim 19, wherein the training stress pattern extractor is configured to perform the categorizing by extracting a level for each of the plurality of cycles from a plurality of levels of a determined parameter, and configured to generate characteristic data based on a determined number of cycles, of the plurality of cycles, that correspond to each of the plurality of levels, so that the characteristic data represents a characteristic of the training stress pattern.

23. The battery life estimation apparatus of claim 22, wherein the determined parameter comprises at least one of an offset, an amplitude, and a period of each of the plurality of cycles.

24. The battery life estimation apparatus of claim 22, wherein the training stress pattern extractor is configured to create a plurality of combination parameters, each representing respective levels for each of a plurality of the determined parameters for a cycle, and configured to perform the categorizing by generating the characteristic data based on a determined number of cycles, of the plurality of cycles, whose determined parameters correspond to each of the plurality of combination parameters.

25. The battery life estimation apparatus of claim 22, wherein the learning parameter determiner is configured to extract the learning parameter by inputting the characteristic data to a predetermined learner.

26. The battery life estimation apparatus of claim 25, further comprising a communication interface, wherein the learning parameter determiner is configured to transmit the extracted learning parameter to an external apparatus using the communication interface.

27. The battery life estimation apparatus of claim 25, further comprising a storage, wherein the learning parameter determiner is configured to store the extracted learning parameter in the storage.

28. A battery life estimation apparatus comprising:

a stress pattern extractor configured to use at least one processing device to generate characterization data that categorizes different stresses of a battery from acquired sensing data of the battery; and

a life estimator configured to use at least one processing device to estimate and output a life of the battery based on the characterization data.

29. The battery life estimation apparatus of claim 28, further comprising a sensor system including a plurality of sensors to measure the sensing data of the battery, the sensing data being real time measurements of physical properties of the battery.

30. The battery life estimation apparatus of claim 28, wherein the life estimator estimates the life of the battery in real time by providing the characteristic data to a learner to which a learning parameter is applied, wherein the learning parameter is previously trained on battery training sensing data of a previous time.

31. A battery life estimation method comprising:

acquiring sensing data for physical properties of a battery;

extracting, using at least one processing device, a stress pattern from the sensing data, the stress pattern representing changes in states of the battery based on stresses applied to the battery and characterized by categorizing different stresses represented in the sensing data; and

estimating a life of the battery based on the categorized stress pattern.

33. A non-transitory computer-readable storage medium comprising computer readable code to cause at least one processing device to perform the method of claim 31.

32. A battery life estimation method comprising:

acquiring training data for physical properties for a battery;

extracting, using at least one processing device, a training stress pattern from the training data, the training stress pattern representing changes in states of the battery based on stresses applied to the battery and characterized by categorizing different stresses represented by the training data; and

determining, using at least one processing device, a learning parameter based on the characterized training stress pattern, the learning parameter being determined for use in estimating a life of the battery.