Optimization of Dephosphorization Parameter in Consteel Electric Arc Furnace Using Rule Set Model

Abstract

The proper process parameter of dephosphorization is a key problem in actual steelmaking production. This article presents a rule set model combining a k‐means method and a decision tree model, which can extract the rule set from the actual production data to optimize the dephosphorization process. The industrial data from Consteel electric arc furnace are selected as the sample data, and then, the phosphorus partition ratio (Lp) is calculated based on the selected data. The k‐means method is used to divide the calculated Lp into two classes (high Lp and low Lp), and meanwhile, through the data statistic, the importance variables, including oxygen consumption, lime weight, dolomite weight, smelting temperature, and scrap ratio (ratio of scrap weight and hot metal weight), are extracted and merged with the classified Lp data into the new data set. Then, the decision tree is used to extract the valuable information to obtain the rule set from new data set. From the rule set, the optimum range of scrap ratio and oxygen consumption under different smelting conditions is obtained, which can guide the adjustment of process parameters in the actual production to enhance the dephosphorization effect.