The operational parameters can be optimized synchronously by inducting and synthesizing similar operational
patterns. For the drawback of the low efficiency of pattern matching cased by huge pattern case, a two-state pattern matching
quickly strategy is proposed. The attribute weights are set by using principal component analysis(PCA) in advance, and the
similarity measure between patterns is also constructed. In order to improve the pattern matching speed, the Cauchy-Schwarz
inequality is introduced in primary pattern matching, which can be used to reduce the computational complexity and narrow
the searching scope for secondary pattern matching. Simulation results based on UCI data sets and actual running data
collected from a copper flash smelting process are given to verify the effectiveness and feasibility.