论文部分内容阅读
冲压成形数值仿真结果中隐含着大量的领域知识。文章将主成分分析与模糊聚类方法应用于基于仿真模型与模拟结果数据的冲压件相似性判别与成形性能判别。通过对油箱冲压成形有限元仿真结果数据进行处理,分析了压边力、拉延筋设置参数、摩擦系数等工艺参数对成形性能的相对重要程度;构造了油箱成形性能的模糊概念,描述其破裂、起皱程度。通过对汽车覆盖件有限元模型数据的分析,对汽车覆盖件进行模糊分类及相似性判别。结果表明,面向有限元仿真结果的数据挖掘技术,可以为冲压成形领域知识发现提供一种有效的新途径。
Numerical simulation results of stamping imply a lot of domain knowledge. In this paper, principal component analysis and fuzzy clustering method are applied to determine the similarity and forming performance of stamping parts based on the simulation model and simulation results. By analyzing the finite element simulation data of the oil tank stamping process, the relative importance of process parameters such as blank holder force, drawbead setting parameters and friction coefficient to the formability was analyzed. The fuzzy concept of the forming performance of the fuel tank was constructed to describe the rupture , Wrinkle level. Through the analysis of the finite element model data of car cover, the fuzzy classification and similarity discrimination of car cover are carried out. The results show that the data mining technology oriented to the finite element simulation results can provide an effective new approach for the knowledge discovery in the field of stamping.