Parameter identification of an electrochemical lithium-ion battery model with convolutional neural network
- Title
- Parameter identification of an electrochemical lithium-ion battery model with convolutional neural network
- Authors
- CHUN, HUI YONG; KIM, JUNGSOO; HAN, SOOHEE
- Date Issued
- 2019-06-11
- Publisher
- IFAC
- Abstract
- Battery is one of the most important energy supplement source for our society. Especially, lithium-ion battery has been actively used in various fields such as mobile devices, electric vehicles, or energy storage system. However, a lithium-ion battery has a few life degradation and safety problems, for example, ignition and explosion. Therefore, it is required to observe the inner states of lithium-ion battery consistently to predict or prevent the problems above. Electrochemical model of lithium-ion battery represents these states thoroughly because it is derived according to the laws of physics. In the electrochemical model, the parameters mean the inner states such as solid particle conductivity, solid particle areas, and solid electrolyte interface layer thickness. In this paper, deep learning algorithm which is a powerful tool to solve complicated problems, is employed to estimate these parameters. Especially, convolutional neural network (CNN) is adopted for low computational burden compared to other deep learning algorithms. The regression results from CNN shows that the parameters could be estimated with relatively high accuracy. © 2019, IFAC (International Federation of Automatic Control) Hosting by Elsevier Ltd. All rights reserved.IFAC (International Federation of Automatic Control) Hosting by Elsevier Ltd. All rights reserved.
- URI
- https://oasis.postech.ac.kr/handle/2014.oak/103136
- Article Type
- Conference
- Citation
- IFAC Workshop on Control of Smart Grid and Renewable Energy Systems(CSGRES) 2019, 2019-06-11
- Files in This Item:
- There are no files associated with this item.
Items in DSpace are protected by copyright, with all rights reserved, unless otherwise indicated.