Methodology of Stamped Metal Parts Retrieval Using Feature Combination and User´s Feedback

Case based reasoning(CBR) is becoming widely used in computer aided process planning system. It requires a mechanism for retrieval of similar parts from a case database which in turn requires a method for case similarity measurement. This paper presents a new approach of stamped metal parts retrieval, which is based on the multi-feature combination and user´s feedback. The stamped metal parts information consists of 3D part model and material which includes material thickness, strength and hardness. The parametric feature relation graph(P-FRG) extract from 3D stamping part model to captures the geometric, topological, and domain specific used to measure the similarity between stamped metal parts. Integrated similarity assessment is achieved by the amalgamation of feature count, feature topology, feature parameter and material. Weight values of Feature and feature component is adjusted based on a statistical analysis of the similarity feature values by user´s feedback.