A knowledge-based system approach for recognition of machining features

This paper describes a knowledge-based system approach for recognition of machining features. Using a machining feature as a positive teaching example, the method converts the boundary information of the machining feature and the user given machining instruction automatically into a set of production rules. With the incorporation of the new rules, the system will be able to recall and apply the learnt machining instruction to other machining features that have shape pattern similar to that of the teaching example. An experimental system to implement the method is presented and the method is explained with illustrations. The method would be useful for predefining new machining features to be recognised.