Similarity Calculation Of 3D CAD Model Based On Ant Colony Algorithm

With the rapid development of manufacturing industry,a lot of three-dimensional(3D)model components are used.So,retrieval technology of3 D CAD model is widely applied.In process of manufacturing and designing,how to retrieve and reuse 3D models efficiently from a large model library is still a research focus.When the scale of model library is large,it is difficult and complex to retrieve a 3D model.These existing model retrieval technologies can not meet needs of users in practice.So,it is very important to explore new approaches of model retrieval and research new technologies of model retrieval.This paper studies CAD model retrieval methods based on greedy strategy and ant colony algorithm.These two methods are well suitable to model retrieval problems and can evaluate similarity degree of two models effectively.The research content of this paper is mainly divided into the following parts:Firstly,this paper introduces research status of CAD model retrieval at home and abroad.At the same time,some new research methods are described.Many classification and optimization methods of CAD model retrieval are studied.Those difficulties in process of CAD model retrieval are given.The existing model library and number of CAD models is very large.So,the efficient of CAD model retrieval is very low.The existing search algorithm has its own limitations and can not adapt to the model retrieval problem.Secondly,feature extraction method of CAD model is studied.The attribute adjacency graph is used to represent 3D model.Model face similarity calculation method is given?According to the difference of edge numbers,shape similarity between two model faces is calculated.At the same time,face similarity matrix is constructed.The adjacency corresponding relation of faces is introduced and thesimilarity of face adjacency corresponding relation is calculated.The process of computing shape similarity of two model faces is described,which is the foundation of model retrieval.Greedy strategy and ant colony algorithm are used to search optimal face matching pairs of two models.At the same time,face matching method,matching flows and experiment process are given.Finally,experimental data of these two retrieval methods are calculated.The advantages and disadvantages are compared between greedy strategy and ant colony search on model retrieval.The accuracy of the ant colony algorithm matching model is higher than the greedy strategy and the greedy strategy can not evaluate the difference of the two complex models.The time complexity of the greedy algorithm is lower than that of the ant colony algorithm.Although the operation efficiency is high,but the matching result is not accurate.