Method and apparatus for automatically estimating remaining useful life (rul) of battery in real time

Patent metadata

Field	Value
Publication number	US-2016259014-A1
Application number	US-201615059521-A
Country	US
Kind code	A1
Filing date	Mar 3, 2016
Priority date	Mar 3, 2015
Publication date	Sep 8, 2016
Grant date	—

Abstract

A method of estimating a remaining useful life (RUL) of a battery includes: identifying a class of data of the battery in real time; determining whether a second level RUL estimation is set for the class; estimating a gross RUL by performing a first level RUL estimation in response to the second level RUL estimation not being set for the class; and estimating a fine RUL of the battery in response to the second level RUL estimation being set for the class.

First claim

What is claimed is:

1 . A method of estimating a remaining useful life (RUL) of a battery, the method comprising:
identifying a class of data of the battery in real time; determining whether a second level RUL estimation is set for the class; estimating a gross RUL by performing a first level RUL estimation in response to the second level RUL estimation not being set for the class; and estimating a fine RUL of the battery in response to the second level RUL estimation being set for the class.

2 . The method of claim 1 , wherein the class is pre-defined based on a number of at least one of charge, discharge, and impedance cycles of the battery.

3 . The method of claim 1 , wherein the identifying of the class comprises:
collecting at least one primary parameter; generating at least one secondary parameter from the at least one primary parameter; generating an optimized set of parameters based on the primary and secondary parameters; generating a real-time artificial intelligence (AI) model specific to the data of the battery based on the optimized set of parameters; comparing the real-time AI model with a reference AI model; and identifying the class in the reference AI model having data matching data in the real-time AI model.

4 . The method of claim 1 , wherein the estimating of the gross RUL comprises measuring a rough estimate of the RUL of the battery.

5 . The method of claim 1 , wherein the estimating of the fine RUL comprises:
collecting battery-specific data identified as belonging to the class; generating a regression model based on the collected battery-specific data; comparing the regression model with a reference regression model representing an optimized data set that represents the class; identifying data in the reference regression model matching data in the regression model; and estimating an RUL representing the identified data in the reference regression model as the fine RUL.

6 . The method of claim 1 , further comprising:
displaying the estimated fine RUL in response to a difference between the fine RUL and an end of life (EOL) of the battery being less than or equal to a threshold value.

7 . A system for estimating a remaining useful life (RUL) of a battery, the system comprising:
an RUL estimator; and a non-volatile memory comprising instructions, wherein the instructions are configured to cause the RUL estimator to:
identify a class of data of the battery in real time;
determine whether a second level RUL estimation is set for the class;
estimate a gross RUL by performing a first level RUL estimation in response to the second level RUL estimation not being set for the class; and
estimate a fine RUL of the battery in response to the second level RUL estimation being set for the class.

8 . The system of claim 7 , wherein the RUL estimator is configured to provide at least one option to pre-define the class based on a number of at least one of charge, discharge, and impedance cycles of the battery.

9 . The system of claim 7 , further comprising a classifier configured to:
collect at least one primary parameter using an input/output (I/O) interface; generate at least one secondary parameter from the at least one primary parameter; generate an optimized set of parameters based on the primary and secondary parameters; generate a real-time artificial intelligence (AI) model specific to the data of the battery based on the optimized set of parameters; compare the real-time AI model with a reference AI model; and identify the class in the reference AI model having data matching data in the real-time AI model, wherein the RUL estimator is configured to identify the class of the data of the battery using the classifier.

10 . The system of claim 7 , further comprising:
a regression analyzer configured to measure a rough estimate of the RUL, wherein the RUL estimator is configured to estimate the gross RUL using the regression analyzer.

11 . The system of claim 7 , wherein, to estimate the fine RUL of the battery, the RUL estimator is configured to:
collect battery-specific data identified as belonging to the class; generate a regression model based on the collected battery-specific data; compare the regression model with a reference regression model representing an optimized data set that represents the class; identify data in the reference regression model matching data in the regression model; and estimate the fine RUL based on the second level RUL estimation by estimating an RUL representing the identified data in the reference regression model as the fine RUL.

12 . The system of claim 7 , further comprising:
a display configured to display the fine RUL in response to a difference between the fine RUL and an end of life (EOL) of the battery being less than or equal to a threshold value.

13 . A remaining useful life (RUL) estimator, comprising:
a processor; a classifier implemented by the processor and configured to identify a class of data of a battery in real time; a state of health (SOH) monitor implemented by the processor and configured to determine whether a second level RUL estimation is set for the class; a regression analyzer implemented by the process and configured to
estimate a gross RUL by performing a first level RUL estimation in response to the second level RUL estimation not being set for the class, and
estimate a fine RUL of the battery in response to the second level RUL estimation being set for the class.

14 . The RUL estimator of claim 13 , wherein the estimating of the gross RUL comprises measuring a rough estimate of the RUL of the battery.

15 . The RUL estimator of claim 13 , wherein the class is pre-defined based on a number of at least one of charge, discharge, and impedance cycles of the battery.

16 . The RUL estimator of claim 13 , wherein the RUL estimator is configured to trigger an alert in response to the estimated fine RUL being greater than an upper threshold value or less than a lower threshold value.

Assignees

• Samsung Electronics Co Ltd

Inventors

• Adagouda Patil Meru
• Seethalakshmy Hariharan Krishnan

Classifications

• G01R31/392 Primary

  Determining battery ageing or deterioration, e.g. state of health · CPC title

• G01R31/367

  Software therefor, e.g. for battery testing using modelling or look-up tables · CPC title

• G01R31/3648

  comprising digital calculation means, e.g. for performing an algorithm · CPC title

• G01R31/3679 Primary

  Physics · mapped topic