A novelty data mining approach for multi-influence factors on billet gas consumption in reheating furnace

Abstract

To systematically and quantitatively analyze the influence factors on billet gas consumption (BGC) in reheating furnace, a novelty data mining approach for multi-influence factors BGC analysis was proposed in this paper. This multi-influence factors data mining model mainly includes four steps: Firstly, the BGC apportionment model was established based on energy apportionment model in reheating furnace; Secondly, the BGC data set could be achieved according to the division of billet sample space (BSS); Thirdly, the data interpolation calculation method of various BSS subsets (BSSSs) was put forward; Lastly, the influence degree analysis method of various factors on BGC was described in detail. Especially, contribution degree model, which could quantitatively describe the influence degree of each factor on BGC, was established. Case study showed that working groups (WGs) should be eliminated because of weak influence on BGC. Then the order of contribution degree on BGC from weak to strong was working shifts (WSs) (1.61%), residence time (9.7%), loading temperature (88.68%). Therefore, residence time and loading temperature should be highlighted in all factors. Finally, some measures and suggestions, which could improve the residence time and loading temperature, were put forward.

Keywords

• Iron and steel industry
• Reheating furnace
• Billet gas consumption
• Multi-influence factors
• Data mining