面向数控加工智能编程的特征自动识别技术

Automatic Feature Recognition Technology for Intelligent Programming of CNC Machining

传统特征识别方法的特征定义融合于特征识别算法中,用户难以根据自身工艺需求自定义目标加工特征。因此,特征识别覆盖范围有限、算法维护周期长且成本高,特征解释的合理性与准确性直接影响工艺设计的效率与质量。为了解决该问题、有效支撑智能化编程模式,本文提出特征组件的概念,将特征的定义转化为组件的定义与聚类,并通过几何属性向量、拓扑邻接矩阵与约束向量定义实现特征统一自定义。同时,给出了自动识别算法流程,结合具体识别实例说明了该方法的实用性与可行性。

The traditional feature recognition method is integrated into feature recognition algorithm, so it is difficult for users to customize target processing features according to their own process requirements. Therefore, the feature recognition coverage is limited, with long algorithm maintenance period and high cost, and the rationality and accuracy of the feature interpretation directly affect the efficiency and quality of the process design. In order to solve this problem and effectively support the intelligent programming mode, this paper proposes the concept of feature components, transforms the definition of features into component definition and clustering, and realizes feature unified customization through geometric attribute vector, topology adjacency matrix and constraint vector definition. At the same time, the automatic identification algorithm flow is given, and the practicality and feasibility of the method are illustrated with specific identification examples.