| 학회 |
한국화학공학회 |
| 학술대회 |
2022년 봄 (04/20 ~ 04/23, 제주국제컨벤션센터) |
| 권호 |
28권 1호, p.120 |
| 발표분야 |
[주제 2] 기계학습 |
| 제목 |
Automatical interpretation of electrochemical impedance spectra data for the equivalent circuit classification on electrochemical systems |
| 초록 |
The rapid development of electrochemical technology has given rise to massive impedimetric and voltammetric datasets. However, in many cases, the dataset was improperly interpreted, which may arise due to the complicated electrochemical system or the compromise between the result and the theoretical model, leading to bias in the interpretation process, especially in the impedimetric result. The electrochemical impedance spectroscopy (EIS) technique plays an essential role in analyzing the properties of the electrode, including their physical and chemical properties. Therefore, the experimenter usually interprets impedimetric results using a searching strategy based on a theoretical basis. It is a time-consuming process, and it is easy to make a mistake. In order to reduce mistakes, we introduced a machine learning (ML) strategy for classifying an EIS equivalent circuit model using a deep neural network. With this approach, the ML model demonstrated a highly accurate classifier with an average AUC of more than 0.95. Intriguingly, the ML model can also classify complicated EIS circuits with significant confidence and an AUC above 0.85. |
| 저자 |
Doonyapisut Dulyawat1, Padmanathan Karthick Kannan2, 김병규3, 정찬화4
|
| 소속 |
1성균관대, 2Department of Chemistry, 3Sungkyunkwan Univ., 4School of Chemical Engineering |
| 키워드 |
에너지 환경(Energy and Environment) |
| E-Mail |
|
| 원문파일 |
초록 보기 |
