DC Capacitor Parameter Estimation Technique for Three-Phase DC/AC Converter Using Deep Learning Methods with Different Frequency Band Inputs
- Authors
- Park, H.-J.; Kwak, Sangshin
- Issue Date
- May-2023
- Publisher
- Korean Institute of Electrical Engineers
- Keywords
- Capacitance; CNN; DC/AC 3-phase converter; Deep learning; DNN; ESR; LSTM; Simple RNN
- Citation
- Journal of Electrical Engineering and Technology, v.18, no.3, pp 1841 - 1850
- Pages
- 10
- Journal Title
- Journal of Electrical Engineering and Technology
- Volume
- 18
- Number
- 3
- Start Page
- 1841
- End Page
- 1850
- URI
- https://scholarworks.bwise.kr/cau/handle/2019.sw.cau/66390
- DOI
- 10.1007/s42835-023-01424-z
- ISSN
- 1975-0102
2093-7423
- Abstract
- In this paper, we estimate the internal state variables of DC/AC 3-phase converter input capacitors according to the input data characteristics of algorithms and compare their performance. There are four deep learning algorithms used in estimation: Deep Neural Network (DNN), Convolution Neural Network (CNN), Simple Recurrent Neural Network (Simple RNN), and Long Short-Term Memory (LSTM). It was selected through frequency characteristic analysis. Deep learning was learned by using the characteristics that the low-frequency component is dominant in capacitance and the mid-frequency component is dominant in Equivalent Series Resistance (ESR). Accordingly, a specific frequency component was used or a broad frequency band including a specific frequency component was used. As a result, there was a suitable algorithm according to the characteristics of the input data. DNN showed excellent estimation performance when specific frequency components were used. On the other hand, when the broad frequency band was used as input data, the performance of CNN was excellent. © 2023, The Author(s) under exclusive licence to The Korean Institute of Electrical Engineers.
- Files in This Item
- There are no files associated with this item.
- Appears in
Collections - College of ICT Engineering > School of Electrical and Electronics Engineering > 1. Journal Articles
![qrcode](https://api.qrserver.com/v1/create-qr-code/?size=55x55&data=https://scholarworks.bwise.kr/cau/handle/2019.sw.cau/66390)
Items in ScholarWorks are protected by copyright, with all rights reserved, unless otherwise indicated.