A multi-source similar subparts based NC process fusion and regrouping approach

Abstract

As a vast number of 3D CAD models associated with NC process are generated each year, taking full advantage of them is an effective way to generate the NC process for query subparts with less time and lower cost. However, there has been little research on how to discover and utilize the valuable information imbedded in the NC process of multiple existing similar subparts. In this paper, a novel multi-source similar subparts-based NC process fusion and regrouping approach is proposed. Firstly, the multi-source similar subparts are described as multi-dimension vectors consisting of feature attributes and NC process. Secondly, the attribute similarities of query feature and similar features are calculated to establish the similarity matrix. Then, based on the multi-source similar features, the gray relational analysis is utilized to mine the association between feature attributes and NC process, and the relation matrix is constructed. Based on the multiplication of similarity matrix and relation matrix, the adaption matrix is calculated to represent the important degree of the NC process of similar features for query feature. Through the weighted sum of adaption values and adjustments, the NC process is obtained for the query subpart. Finally, based on the feature interactions in the query subpart, the NC process of the query subpart is regrouped to meet the requirements of NC machining.