BROWSE
Detail
Conference Paper
Deep Learning for Forecasting Production Quantity of Ferro-alloy in Electric Arc Furnaces (EAF)
Cited - time in 
Share 
Share
- Authors
- O SangWon, Young Seog Yoon, Kwangroh Park
- Issue Date
- 2019-08
- Citation
- International Conference on Big Data Applications and Services (BigDAS) 2019, pp.1-9
- Language
- English
- Type
- Conference Paper
- Abstract
- Large fluctuations in production quantity of Ferro-alloy manufactured in an electric arc furnace (EAF) prevents manufactures to establish a reliable manufacturing and operating plan. However, little is known about the factors and their effects on Ferro-alloy production in EAF. Moreover, the nature of a furnace (extremely high temperature in a fully-closed space) reinforces the uncertainty. To fill out the gap, this study applied deep learning models to forecasting production quantity of Ferro-alloy. Specifically, we employed recursive neural network (RNN), Long Short Term Memory (LSTM), and Gated Recurrent Unit (GRU) with two-year data of daily input power, daily pressing depth of three electrodes, the number of daily productions, and a dummy variable of consistency in producing same product. Prediction performance of deep learning models indicate that deep learning might provide a useful guideline for establishing manufacturing plan.
- KSP Keywords
- Closed space, Electric arc furnace, Gated recurrent unit, High Temperature, Input power, Long-short term memory(LSTM), Recursive neural network, deep learning(DL), deep learning models, dummy variable, prediction performance