Research On Feature Recognition Method Of Sheet Metal Parts Based On Graphs And Rules

Implementation of automatic machining process planning for sheet metal is of great significance to improve the production efficiency of sheet metal parts.Automatic process planning depends on the extraction of process features from CAD part by CAPP system,and feature recognition technology is one of the most effective methods to solve process feature extraction.However,current feature recognition technology still has no very effective solution for sheet metal feature recognition.So this thesis which takes sheet metal parts as the research object carries out research on feature recognition algorithm based on graphs and rules,and realizes the extraction of main process features in sheet metal parts.The research work is as follows:By studying the machining process and technology of sheet metal parts,and analysising the computer internal expression and topological adjacency of sheet metal part’s three-dimensional model,a method of classifying the characteristics of sheet metal parts is summarized.Based on all above,the basic attribute adjacency graph is constructed to provide a theoretical basis for accurate classification and extraction of sheet metal features.Traditional attribute adjacency graph mainly expresses the topological adjacency relationship of the part model,but it can not accurately represent the specific process feature of sheet metal.For this reason,an algorithm which constructs the extended attribute adjacency graph of sheet metal is proposed to provide the basis for the decomposition and extraction of sheet metal process features.Current graph-based feature recognition algorithm can not recognize the bending feature and composite feature of sheet metal.For this reason,sheet metal convex decomposition and rule-based feature recognition algorithms are proposed.The extended attribute adjacency graph is convex decomposed by the convex face and convex edge.Convex face decomposition is used to extract the blend features and bending features of the sheet metal parts.Convex edge decomposition is used to extract the notch features and simple inner contour features of the sheet metal parts.Combined with the insufficient decomposition and excessive decomposition of composite inner contour features and array features in the convex decomposition,the subgraphs are processed and reconstructed based on rules.And process features of sheet metal are recognized by heuristic rules.Based on the sheet metal convex decomposition feature recognition algorithm and rulebased feature recognition algorithm,the feature recognition module is developed on Inte3 D software platform,and typical sheet metal parts are taken as examples to verify the feature recognition algorithm.The results show that the algorithm is feasible and effective.