HEAT LOAD CONTROL OF BLAST FURNACE WALL USING STATISTICAL OPTIMIZATION TECHNIQUES

摘要

A statistical optimization method is proposed to maintain stable operation in a blast furnace with stave coolers. The objective of this study is to find the operating condition which maximizes productivity as well as minimizes the heat load on the stave coolers. In this method, we first analyzed historical data and calculated the heat flux of stave coolers and the burden distribution by mathematical models. In this approach, we determined the variables to optimize and manipulated variables to adjust these response variables. Next, we collected the necessary data for the data variables after preprocessing such as removal of noise and outlier. Finally, we constructed empirical models describing the patterns of response variables in terms of manipulated variables by partial least squares regression method, and then these models were used as objective functions for entire optimization problem. By using appropriate optimization algorithm, we determined the compromised optimum operating condition considering all response variables from the past operating conditions.