节点文献

中文自然语言驱动的自动几何建模方法

Automatic geometrical modeling method driven by Chinese natural language

  • 推荐 CAJ下载
  • PDF下载
  • 不支持迅雷等下载工具,请取消加速工具后下载。

【作者】 王振伟田野陈丰刘云杰

【Author】 WANG Zhen-wei;TIAN Ye;CHEN Feng;LIU Yun-jie;School of Astronautics and Aeronautics,University of Electronic Science & Technology of China;Sichuan Special Equipment Institute;

【机构】 电子科技大学航空航天学院四川出入境检验检疫局

【摘要】 针对目前几何建模的人机交互能力不足、设计自动化水平不高等问题,提出中文自然语言驱动的自动几何建模方法。首先给出中文自然语言驱动的自动几何建模方法框架,对几何建模意图的自然语言描述与理解方法进行论述,建立了几何建模意图的自然语言描述规则和几何建模意图表达范式,提出了基于几何要素模式匹配的自然语言规范处理算法;然后阐述了几何要素的数据表达和映射方法,论述了几何要素表达范式的自动推理与建模数据模式识别,并对几何要素的STEP自动映射进行分析,给出几何要素冗余信息的处理过程,从而实现中文自然语言驱动的自动几何建模。通过开发相关的原型系统进行实验研究,验证了所提方法的合理性和可行性。

【Abstract】 Aiming at the deficiency of human-computer interaction’s ability and design automation level for geometrical modeling,the automatic geometrical modeling method driven by Chinese natural language was proposed.The automatic geometrical modeling framework driven by Chinese natural language was given.The description and understanding methods for geometrical intention natural language were discussed,and the corresponding representations were constructed.The standard processing algorithm for natural language based on geometrical element matching was proposed,and the standard description of geometrical modeling intention’s natural language was realized. The data expression and mapping method of geometrical modeling elements were analyzed,and the mapping mechanism of modeling intension was constructed.The STandard for the Exchange of Product model data(STEP)automatic mapping of geometrical element was elaborated and the processing procedure of the redundant information was presented,which could realize the automatic geometrical modeling of Chinese natural language driven.The relevant protosystem was developed,and the result showed the feasibility and the rationality of the proposed method.

【基金】 国家自然科学基金资助项目(51005040);中国博士后科学基金资助项目(20100470076)~~
  • 【文献出处】 计算机集成制造系统 ,Computer Integrated Manufacturing Systems , 编辑部邮箱 ,2014年02期
  • 【分类号】TP391.7
  • 【被引频次】5
  • 【下载频次】106
节点文献中: 

本文链接的文献网络图示:

本文的引文网络