In this research work, deep machine learning based methods together with a novel data augmentation are developed for predicting flicker, voltage dip, harmonics and interharmonics originating from highly time-varying electric arc furnace (EAF) currents and voltage. The aim with the prediction is to counteract both the response time delays and reaction time delays of active power filters (APFs) specifically designed for electric arc furnaces (EAF). Multiple synchronous Reference frame (MSRF) analysis is used to decompose the frequency components of the EAF current and voltage waveforms into dqo components. Then using low pass filters and prediction of the future values of these dqo components, reference signals for APFs are generated. Three different methods have been developed. In two of them, a low pass Butterworth filter is used together with a linear FIR based prediction or long short term memory network (LSTM) for prediction. In the third method, a deep convolutional neural network (CNN) combined with and LSTM network is used to filter and predict at the same time. For a 40 ms prediction horizon, the proposed methods provide 2.06, 0.31, 0.99 prediction errors of the dqo components for the Butterworth and linear prediction, Butterworth and LSTM and CNN with LSTM, respectively. The error of the predicted reconstructed waveforms of flicker, harmonics, and interharmonics resulted in 8.5, 1.90, and 3.2 reconstruction errors for the above-mentioned methods.
Yayın Adı (dc.title) | Predictive Compensation of EAF Flicker, Voltage Dips Harmonics and Interharmonics Using Deep Learning |
Yazar/lar (dc.contributor.yazarlar) | Ebrahim Balouji, Özgül Salor, Tomas McKelvey |
Yayın Türü (dc.type) | Konferans Bildirisi |
Dil (dc.language) | İngilizce |
Yayımlanma Yılı (dc.date.issued) | 2021 |
Ulusal/Uluslararası (dc.identifier.ulusaluluslararasi) | Uluslararası |
Kaynak (dc.relation.journal) | IEEE Industry Applications Society Annual Meeting |
Kaynak Adı Ek Bilgi / Konferans Bilgisi (dc.identifier.kaynakadiekbilgi) | 2021 IEEE Industry Applications Society Annual Meeting. 10-14 October 2021, Vancouver, BC, Canada |
ISSN/ISBN (dc.identifier.issn) | ISSN: 0197-2618 |
Yayıncı (dc.publisher) | IEEE Xplore Digital Librar, USA |
Veri Tabanları (dc.contributor.veritaban) | Web of Science Core Collection |
Veri Tabanları (dc.contributor.veritaban) | IEEE Xplore |
Veri Tabanları (dc.contributor.veritaban) | Scopus |
İndex Türü (dc.identifier.index) | CPCI-S |
İndex Türü (dc.identifier.index) | Scopus |
Özet (dc.description.abstract) | In this research work, deep machine learning based methods together with a novel data augmentation are developed for predicting flicker, voltage dip, harmonics and interharmonics originating from highly time-varying electric arc furnace (EAF) currents and voltage. The aim with the prediction is to counteract both the response time delays and reaction time delays of active power filters (APFs) specifically designed for electric arc furnaces (EAF). Multiple synchronous Reference frame (MSRF) analysis is used to decompose the frequency components of the EAF current and voltage waveforms into dqo components. Then using low pass filters and prediction of the future values of these dqo components, reference signals for APFs are generated. Three different methods have been developed. In two of them, a low pass Butterworth filter is used together with a linear FIR based prediction or long short term memory network (LSTM) for prediction. In the third method, a deep convolutional neural network (CNN) combined with and LSTM network is used to filter and predict at the same time. For a 40 ms prediction horizon, the proposed methods provide 2.06, 0.31, 0.99 prediction errors of the dqo components for the Butterworth and linear prediction, Butterworth and LSTM and CNN with LSTM, respectively. The error of the predicted reconstructed waveforms of flicker, harmonics, and interharmonics resulted in 8.5, 1.90, and 3.2 reconstruction errors for the above-mentioned methods. |
URL (dc.rights) | https://ieeexplore.ieee.org/document/9677400 |
DOI (dc.identifier.doi) | 10.1109/IAS48185.2021.9677400 |
Fakültesi / Enstitütü (dc.identifier.fakulte) | Mühendislik Fakültesi |
Bölümü (dc.identifier.bolum) | Elektrik-Elektronik Mühendisliği Bölümü |
Kurumdaki Yazar/lar (dc.contributor.author) | Özgül SALOR-DURNA |
Kayıt No (dc.identifier.kayitno) | BL5EA32DF9 |
Kayıt Giriş Tarihi (dc.date.available) | 2022-09-08 |
Not (Yayımlanma Yılı) (dc.identifier.notyayinyili) | 2021 |
Wos No (dc.identifier.wos) | WOS:000821959500204 |
Konu Başlıkları (dc.subject) | active power filter (APF) |
Konu Başlıkları (dc.subject) | butterworth filter |
Konu Başlıkları (dc.subject) | convolutional neural networks (CNN) |
Konu Başlıkları (dc.subject) | deep learning (DL) |
Konu Başlıkları (dc.subject) | long short-term memory (LSTM) |
Konu Başlıkları (dc.subject) | linear prediction |
Konu Başlıkları (dc.subject) | multiple synchronous reference frame (MSRF) |
Konu Başlıkları (dc.subject) | electric arc furnace (EAF) |
Konu Başlıkları (dc.subject) | flicker |
Konu Başlıkları (dc.subject) | harmonics |
Konu Başlıkları (dc.subject) | interharmonics |
Konu Başlıkları (dc.subject) | voltage dip |