Battery diagnosing apparatus and method

Patent metadata

Field	Value
Publication number	US-2022413057-A1
Application number	US-202217716461-A
Country	US
Kind code	A1
Filing date	Apr 8, 2022
Priority date	Feb 9, 2021
Publication date	Dec 29, 2022
Grant date	—

Abstract

A battery diagnosing apparatus includes: a characteristic value extracting unit for extracting a plurality of characteristic values for each of a plurality of batteries; a dimension reducing unit for reducing a dimension of a characteristic value profile representing a distribution of the plurality of batteries using a predetermined algorithm based on the plurality of characteristic values extracted by the characteristic value extracting unit; and a state diagnosing unit for detecting an outlier in the characteristic value profile whose dimension is reduced by the dimension reducing unit, and diagnosing a state of each of the plurality of batteries based on the detected outlier.

First claim

1 . A battery diagnosing apparatus, comprising:
a battery management system (BMS) including:
a characteristic value extracting unit configured to extract a plurality of characteristic values for each of a plurality of batteries;
a dimension reducing unit configured to reduce a dimension of a characteristic value profile representing a distribution of the plurality of batteries using a predetermined algorithm based on the plurality of characteristic values extracted by the characteristic value extracting unit; and
a state diagnosing unit configured to detect an outlier in the characteristic value profile whose dimension is reduced by the dimension reducing unit, and diagnose a state of each of the plurality of batteries based on the detected outlier.

2 . The battery diagnosing apparatus according to claim 1 ,
wherein the state diagnosing unit is configured to set a criterion value in the characteristic value profile whose dimension is reduced, calculate a weight and a Euclidean distance for each of the plurality of batteries based on the set criterion value; and detect the outlier based on the weight and the Euclidean distance.

3 . The battery diagnosing apparatus according to claim 2 ,
wherein the state diagnosing unit is configured to calculate the weight for the plurality of batteries based on compactness of the plurality of batteries with respect to the set criterion value.

4 . The battery diagnosing apparatus according to claim 2 ,
wherein the state diagnosing unit is configured to calculate an outlier index by multiplying the Euclidean distance and the weight corresponding to each of the plurality of batteries, and detect the outlier based on the calculated outlier index and a preset threshold value.

5 . The battery diagnosing apparatus according to claim 4 ,
wherein the state diagnosing unit is configured to detect a battery among the plurality of batteries of which the calculated outlier index is equal to or greater than the threshold value as the outlier.

6 . The battery diagnosing apparatus according to claim 2 ,
wherein the state diagnosing unit is configured to classify the plurality of batteries into at least one group in the characteristic value profile, and detect the outlier for each group.

7 . The battery diagnosing apparatus according to claim 1 ,
wherein the characteristic value extracting unit is configured to extract the plurality of characteristic values having orthogonality with each other for each of the plurality of batteries.

8 . The battery diagnosing apparatus according to claim 1 ,
wherein the dimension reducing unit is configured to set perplexity, which is one of parameters of the predetermined algorithm, to be less than the number of the plurality of batteries.

9 . The battery diagnosing apparatus according to claim 1 ,
wherein the state diagnosing unit is configured to normalize the characteristic value profile whose dimension is reduced, and detect the outlier from the normalized characteristic value profile.

10 . A battery pack, comprising the battery diagnosing apparatus according to claim 1 .

11 . An energy storage system, comprising the battery diagnosing apparatus according to claim 1 .

12 . A battery diagnosing method, comprising:
a characteristic value extracting step of extracting, via a battery management system (BMS) a plurality of characteristic values for each of a plurality of batteries; a dimension reducing step of reducing, via the BMS, a dimension of a characteristic value profile representing a distribution of the plurality of batteries using a predetermined algorithm based on the plurality of characteristic values extracted in the characteristic value extracting step; an outlier detecting step of detecting, via the BMS, an outlier in the characteristic value profile whose dimension is reduced in the dimension reducing step; and a state diagnosing step of diagnosing, via the BMS, a state of each of the plurality of batteries based on the outlier detected in the outlier detecting step.

Assignees

• Lg Energy Solution Ltd

Inventors

• Song Yee-Gahng

Classifications

• G01R31/3842 Primary

  combining voltage and current measurements · CPC title

• G01R19/16542

  for batteries (charge condition monitoring in G01R31/36) · CPC title

• G01R31/396 Primary

  Acquisition or processing of data for testing or for monitoring individual cells or groups of cells within a battery · CPC title

• H01M10/482

  for several batteries or cells simultaneously or sequentially · CPC title

• Y02E60/10

  Energy storage using batteries · CPC title