摘要
基于粗糙集理论的数据挖掘技术 ,研究了面向有限元仿真结果的知识自动获取方法 .以压边力、摩擦系数、厚向各向异性系数、凹模圆角半径、板料厚度等作为条件属性 ,成形后零件两种主要的成形缺陷——破裂和起皱作为目标属性 ,分析了圆筒件拉深成形中材料性能、工艺参数对成形性能的影响程度 ,提炼出对加工工艺和模具设计有指导意义的产生式规则 .结果表明 ,面向有限元仿真结果的数据挖掘技术为塑性成形领域知识发现提供了一条有效的新途径 .
The method of knowledge acquisition from finite element simulation results is studied with data mining technology based on rough set theory. The cylindrical cup drawing process is selected as an example, taking blankholder force, friction coefficient, normal anisotropic parameter and blank thickness as conditional attributes, two main defects after forming - crack and wrinkle as object attributes. The influence extent that different material properties and process parameters exert on the formability is analyzed. Many instructive 'what if' rules are abstracted, which are helpful to the process planning and die design. The results show that the data mining technology oriented to FEM results provides an effectively novel method of knowledge acquisition in the field of plastic forming.
出处
《上海交通大学学报》
EI
CAS
CSCD
北大核心
2004年第7期1065-1068,共4页
Journal of Shanghai Jiaotong University
基金
上海市科技启明星计划跟踪项目 ( 0 1QMH14 11)
关键词
塑性成形
数值模拟
粗糙集
知识获取
数据挖掘
Data mining
Finite element method
Knowledge acquisition
Plastics forming
Rough set theory
