Abstract: Aiming at the machining feature recognition problem of engine cylinder blocks,a hybrid machining feature recognition method based on graph decomposition and feature factors cluster was proposed.The machining surfaces were obtained from 3D models in STEP AP203 format,and the extended attributed adjacency graph was constructed.The extended attributed adjacency graph was decomposed into the minimum attributed adjacency graphs which could be matched with predefined feature elements to obtain the characteristic factors.To meet the machining requirements,the feature factors were clustered based on machining rules and the composite machining feature were obtained.A case study was carried out to demonstrate the feasibility and effectiveness of the proposed method.The experimental results illustrated that the method could effectively solve some problems of intersecting features recognition and NP-hard feature explanation for cylinder blocks.

Key words: cylinder block, machining feature recognition, attributed adjacency graph, composite machining feature

摘要： 针对发动机缸体零件的加工特征识别问题,提出一种基于图分解与特征因子聚类的混合式加工特征识别方法。对零件STEP AP203模型数据进行预处理,获得特征加工面,构建零件加工面扩展属性邻接图;通过分解获得最小属性邻接图,并与预定义的特征元匹配,获得特征因子;为满足特征识别结果工艺性的要求,对特征因子聚类,构造复合加工特征。通过具体实例验证了所提方法的有效性和可行性,表明该方法能有效解决现有缸体零件相交特征识别困难和特征解释NP问题。

关键词: 缸体类零件, 加工特征识别, 属性邻接图, 复合加工特征