Research And System Implementation Of 3D Model Shape Feature Recognition Algorithm Under NX Platform

With the introduction of the concept of Industry 4.0 and the issuance of the "Made in China 2025" strategic document,the process of the fourth industrial revolution has continued to deepen.And intelligent manufacturing has become a research hotspot in the industry in recent years.In the field of intelligent manufacturing,industrial modeling software plays a vital role in the digitalization,networking and intelligence of industrial manufacturing.In industrial modeling software,the identification and application of shape features of part models is the key point for realizing intelligent design and manufacturing.Therefore,this thesis has carried out in-depth research on shape feature recognition technology,and designed a relatively complete solutions for the problem of quick selection of similar regions.Based on the NX software platform,with the help of the NX secondary development framework,this thesis developed a similar region search subsystem for quick selection of similar regions.First,the specific requirements of the similar region search system are analyzed.What the system needs to implement is to allow the user to select a shape region,a group of connected faces,and search in the part model for other regions that are similar in topology and shape to the selected region.At the same time,the selected similar regions need to be in one-to-one correspondence with the region selected by the user,so as to simplify the operation of the repeated processing design specific to the faces.Second,feature extraction and reconstruction are performed on the shape region.In order to ensure the exact one-to-one correspondence of faces,an attribute adjacency graph is constructed according to the adjacency relationship of the faces by taking the region faces as nodes.With the help of NX interfaces to extract the feature information of region face and region edge,the parameterized shape region attribute adjacency graph model is reconstructed from the solid model.Third,the shape feature recognition algorithm of the core of the system is designed and researched.For the attribute adjacency graph of the shape region,the subgraph isomorphism algorithm is used to ensure that the topological structure of the candidate region and selected region is consistent.And the accuracy of the surface mapping relationship is ensured with the help of region face shape feature information and spatial position information.Through subgraph isomorphism and face mapping algorithm,similar regions are searched and identified,and the mapping results of similar regions required by the system are obtained.Finally,a reuse method of shape region design knowledge is designed.The attribute adjacency graph model of the shape region is exported to an XML file and saved,and the shape region model can be reconstructed from the file and imported into the system,and then be used for similar region search.This enables knowledge reuse of shape region.On the basis of the above solutions,and with the help of the NX secondary development framework,a similar region search subsystem is designed and implemented.The subsystem is embedded in the NX software system,and the similar area search system is tested and analyzed with the help of the NX automatic test tool,which verifies the effectiveness of the system.