基于部分先验识别的B-Rep模型全局对称性识别方法

摘要

B-Rep模型对称性识别能有效地提高CAE分析效率.针对现有方法通过识别B-Rep模型所有局部对称性来确定全局对称性,效率较低的问题,提出一种基于部分先验识别的B-Rep模型全局对称性识别方法.该方法以一致面集为对称性识别单元,首先利用先验知识快速确定部分局部对称性,然后通过几何推理逐步缩小候选对称性识别范围,最终快速确定B-Rep模型全局对称性.实验结果证明,文中方法鲁棒性高且识别速度快.%Detecting symmetries of B-Rep models can greatly improve the efficiency of CAE analysis.The efficiency of the existing methods is quite low because all the local symmetries of B-Rep models must be detected in advance.A method,based on partial prior recognition,is proposed to detect the global symmetries for B-Rep models.The congruent face set is taken as the unit for symmetry detection.Firstly,the priori knowledge is used to quickly determine the part of the local symmetries.Secondly,the candidate symmetry identification range is shrunk gradually by geometric reasoning.Consequently,the global symmetries of the B-Rep model are determined.Experimental results show that the proposed method is robust and fast.