Using Generalized Iterated Function Systems to Model BF Hot Metal Silicon Content Sequences

A new iterated function system (IFS) model named piecewise self-affine fractal is developed to fit silicon content sequences that collected online from No.7 blast furnace (BF) at Handan Iron and Steel Group Co.. Such algorithm consists of three steps: step 1 selects any [Si] time series to be sample space. And then proves that [Si] time series is a long-term negative association fractal time series using rescaled range analysis; step 2 establishes the piecewise self-affine fractal model which is a generalization of the linear fractal interpolation; step 3 determines the model parameters such as interpolation points, the contraction factors and the addresses associated with each section of the function. The rate of hit shot of silicon content is more than 84% in [Si] plusmn 0.1% range using such new model