Abstract This paper presents an innovative method for recognizing manufacturing features from computer aided design (CAD) part models. The proposed method is to integrate the graph-based approach and the volume approach, in order to combine the positive aspects of both strategies. Feature edge sequence, a new graph-based feature recognition approach, is used to recognize and extract surface features from the part design data. The extracted features are then clustered into the machining volumes by the volume-based approach. The main drawback of conventional feature recognition systems is their limitations in handling feature interactions and arbitrary shape features. In the proposed system, the graph-based method is able to recognize interacting features, the volume-based approach can generate alternative interpretations of machining volumes and an artificial intelligence (AI)-based algorithm, established with a neural network, is used to handle the arbitrary features.
[1]
C. R. Liu,et al.
Analysis on machined feature recognition techniques based on B-rep
,
1996,
Comput. Aided Des..
[2]
Jami J. Shah,et al.
Automatic recognition of interacting machining features based on minimal condition subgraph
,
1998,
Comput. Aided Des..
[3]
T. C. Chang,et al.
Graph-based heuristics for recognition of machined features from a 3D solid model
,
1988
.
[4]
Willem F. Bronsvoort,et al.
Feature modelling and conversion: key concepts to concurrent engineering
,
1993
.
[5]
Jami J. Shah.
Assessment of features technology
,
1991,
Comput. Aided Des..
[6]
Rangasami L. Kashyap,et al.
Geometric Reasoning for Recognition of Three-Dimensional Object Features
,
1990,
IEEE Trans. Pattern Anal. Mach. Intell..