Where academic tradition
meets the exciting future

Analysis of Blast Furnace Time Series Data with ANFIS

Kaj-Mikael Björk, Markus Holopainen, Robin Wikström, Henrik Saxen, Christer Carlsson, Miika Sihvonen, Analysis of Blast Furnace Time Series Data with ANFIS. TUCS Technical Reports 1094, TUCS, 2013.

Abstract:

Analyzing processes that cannot fully be measured is necessary for better decision making. In this paper we analyze some blast furnace data with the help of ANFIS (Artificial Neural Fuzzy Inference System). The entire process is described, from the application, preprocessing the data and the analysis made. Also the results from the best ANFIS models are compared to results with more traditional ARMA-model theory (as a benchmark). Special features in this industrial case study, the blast furnace process, includes by exhibiting high levels of noise and complexity.

Files:

Full publication in PDF-format

BibTeX entry:

@TECHREPORT{tBjHoWiSaCaSi13a,
  title = {Analysis of Blast Furnace Time Series Data with ANFIS},
  author = {Björk, Kaj-Mikael and Holopainen, Markus and Wikström, Robin and Saxen, Henrik and Carlsson, Christer and Sihvonen, Miika},
  number = {1094},
  series = {TUCS Technical Reports},
  publisher = {TUCS},
  year = {2013},
  ISBN = {978-952-12-2986-2},
}

Belongs to TUCS Research Unit(s): Institute for Advanced Management Systems Research (IAMSR)

Edit publication