以製程需求表達法辨認切削特徵

Abstract

本研究旨在應用製程需求表達法(process requirement modeling )來解決電腦製輔助規劃(CAPP)中的自動辨識特徵(machined feature recognition)問題。過去研究對加工特徵自動辨識，通常認為是一個廣義的圖形比對問題，這種方法在特徵辨識上相當受限於特徵的定義。本研究以Lee (2017)的擴充式製程需求表達法來求解特徵辨識問題，該表達法是Wu & Liu (1991)年所發展的製程需求表達法的延伸，所謂製程需求表達法是描述每一個加工面可加工刀具的範疇，若一個集合的相鄰加工面可用同一刀具加工，則此集合的加工面可視為是一個加工特徵。與過去研究相比，本研究所提的方法優點有二： (1) 簡化事前定義加工特徵，(2) 解決交錯特徵問題。

This research attempts to solve the machined feature recognition problem in the area of computer aided process planning. Prior studies considered the machined feature recognition problem as a generalized graph-matching problem. Due to the requirement of defining a machined feature as a graph in advance, the prior approach is limited in the scope of recognizable machined features because such a machined feature definition paradigm is enumerative and can never be exhaustive. In this research, we use the enhanced process requirement modeling developed by Lee (2017) to solve the machined feature recognition problem. The enhanced process requirement modeling, an extension of Wu & Liu (1991), models a machined face by the scope of feasible cutters that can machine the face. Then, a set of connected machined faces that can be machined by a common cutter is called a machined feature. Compared to prior studies, the proposed approach is distinguished in two perspectives: (1) we don’t need to define machined features in advance in an enumerative manner, and (2) we can effectively recognize interacting machined features that were difficult to recognize by prior studies.