论文部分内容阅读
以山东莱钢1号高炉和山西临钢6号高炉在线采集的铁水含硅量([Si])的时间序列为样本,利用重标级差分析(R/S)和盒维数计算方法,得出两座高炉[Si]时间序列的Hurst指数H分别约为0.257和0.224,盒维数DB分别约为1.712和1.724,完成了对[Si]值波动复杂程度的定量估计,证明两者的时间序列均为长程负相关的分形时间序列.随后根据分形迭代函数的理论与方法,确定相关参数,迭代生成模拟的[Si]时间序列,拟合效果较好.
The time series of silicon content ([Si]) in hot metal collected from No. 1 blast furnace at Shandong Laigang and No. 6 blast furnace at Shanxi Linfen was taken as a sample. Using the method of differential grade analysis (R / S) and box dimension calculation, The Hurst exponents H of the two blast furnace [Si] time series are about 0.257 and 0.224, respectively. The box dimensions DB are about 1.712 and 1.724, respectively. The quantitative estimation of [Si] volatility complexity is completed, The sequences are both long-range and negative-correlated fractal time series.According to the theory and method of fractal iterative function, the relevant parameters are determined and the simulated [Si] time series are iteratively generated, which has a good fitting effect.