基于粒子群优化的三维CAD模型相似性评价

Improvement of Similarity Evaluation of 3D CAD Model Based on Particle Swarm Optimization

为了准确度量CAD模型的形状差异以更好地实现重用,提出了基于粒子群优化的CAD模型相似性评价算法。算法首先模型面的边数构造评价矩阵,然后通过粒子群算法搜索两模型的面最优匹配序列,最后提取并计算面相似度,完成对整体模型的相似度计算。结果表明,与已有算法相比,所提算法可以更准确地描述三维模型的相似性,有助于典型结构的准确挖掘重用。

In order to accurately measure the shape difference of the CAD model to better realize multi-granularity and intelligent reuse requirements,a similarity evaluation algorithm for 3D CAD model based on particle swarm optimization is proposed.The similarity evaluation matrix from the number of edges that make up each face of the two models to be evaluated is constructed firstly,and then using that matrix as the descriptive body,the surface optimal matching sequence of the two models is searched through the particle swarm optimization algorithm.Finally,based on the searched optimal surface sequence,the surface similarity is extracted and calculated,and then the overall similarity evaluation of the two models is realized.Experimental results based on the general ESB model library show that,compared with the existing algorithms,the proposed algorithm can better evaluate the similarity of 3D models,and facilitate accurate mining of typical structures to support local structural level design reuse of models.