Quantitative association rule mining for blast furnace production data

Abstract

Blast furnace (BF) ironmaking process is a typical complex nonlinear industrial process. Aiming at the problem that the relationship between the operating parameters and the main production indicators in BF ironmaking process mainly depends on the subjective experience of the specialized operators and experts, and is difficult to be inherited and studied later, this paper introduces data mining technology to analyze the large amount of data produced during the BF ironmaking process and dig out the inherent relationship contained in the data. In order to mine more valuable and interested association relationship, an improved Apriori algorithm is proposed to obtain the quantitative relation between operating parameters and production indicators for the characteristics of BF production data. The experimental results show that the proposed Apriori algorithm can obtain the objective and effective information implied in the production data. The mining rules provide a theoretical basis for blast furnace operation decision and production optimization.