摘要
特征识别是语义特征造型系统面临的一个难题。尤其是当遇到复杂的拓扑结构或是较多的特征数量时,如何提高特征识别的效率以及采取何种识别方法,成为该技术领域的热门课题之一。通过运用BAM神经网络检测技术,提出了一种全新的特征识别策略,有效解决了CAD/CAM系统的性能瓶颈问题。
Recognition of feature is an important topic in feature-based system. The ways used at large about feature checking can not be use for product models with complicated to-pology and lots of features. Improving feature checking efficiency is one of the most challenge questions in the researching on features modeling technology. A new feature strategy was put forward by BAM neural network checking technology, and this method solved effectively the problem about CAD/ CAM capability limitation.
出处
《廊坊师范学院学报（自然科学版）》
2011年第6期24-26,共3页
Journal of Langfang Normal University(Natural Science Edition)
关键词
语义特征造型
BAM神经网络
矢量正交化
多重训练方法
Semantic feature modeling
BAMneural networks
vectors orthogonalization
muhi-training method
