Large Scale Database-based Online Modeling using ICA of Visualized Process Data for Blast Furnace Operation

The large scale database-based online modeling, called LOM, is a type of just-in-time modeling for blast furnace. The database of LOM is so far built by quantizing directly measurement process data. Recently it has been shown that the image data generated by visualizing shaft pressure and stave temperature is very useful for blast furnace operation and guidance. In this paper we try to extend LOM to the one incorporated with the visualized process data. First we extract features of the visualized process data by using independent component analysis (ICA), and add the features (independent components) of the visualized process data, as process data, to the database of LOM. Prediction performance of the extended LOM is illustrated by using real process data