This paper presents an intelligent feature recognition method for STEP-NC-compliant manufacturing based on artificial bee colony (ABC) algorithm and back propagation (BP) neural network. In the method, after extracting the geometric and topological information from its STEP AP203 neutral file, the minimum subgraphs of a part are firstly constructed based on the concavity and convexity judgment algorithm. Then, an improved BP neural network used to STEP-NC-compliant manufacturing feature recognition is proposed with the combination with ABC algorithm. Finally, the STEP-NC-compliant manufacturing features in the part are recognized accurately and efficiently after the information data from the minimum subgraphs of the part is input into the improved BP neural network. At the end, it has been concluded by case study that the proposed method is effective and feasible.

本文提出了一种基于人工蜂群（ABC）算法和反向传播（BP）神经网络的STEP-NC兼容制造智能特征识别方法。在该方法中，从其STEP AP203中性文件中提取几何和拓扑信息之后，首先根据凹凸判断算法构造零件的最小子图。然后，结合ABC算法，提出了一种改进的BP神经网络，用于与STEP-NC兼容的制造特征识别。最后，将来自零件最小子图的信息数据输入到改进的BP神经网络后，可以准确，有效地识别零件中符合STEP-NC的制造特征。在最后，