摘要
在利用相似模式的归纳与综合实现铜闪速熔炼过程多操作参数同步优化的过程中,针对模式库庞大引起相似模式发现效率低的问题,提出了两级模式快速匹配方法.该方法采用主元分析实现属性权重设定,通过构造模式间相似性度量准则,并在初级匹配准则中引入柯西不等式,以减少计算复杂度,快速缩小次级匹配的搜索范围,提高模式匹配速度.UCI数据集仿真和实际生产数据应用结果验证了该方法的有效性和可行性.
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.
出处
《控制与决策》
EI
CSCD
北大核心
2013年第1期120-124,130,共6页
Control and Decision
基金
国家自然科学基金项目(61104135
60904077)
国家863计划项目(2009AA04Z137)
中央科研基本业务费中南大学国家杰青培育专项基金项目(2011JQ009)
湖南省自然科学基金项目(09JJ3122)
关键词
铜闪速熔炼过程
操作模式
模式匹配
柯西不等式
主元分析
copper flash smelting process: operational pattern: pattern matching
Cauchy-Schwarz inequality, principalcomponent analysis