东北大学学报（自然科学版）

东北大学学报:自然科学版 ›› 2015, Vol. 36 ›› Issue (2): 284-288.DOI: 10.12068/j.issn.1005-3026.2015.02.028

• 资源与土木工程 • 上一篇    下一篇

基于萤火虫算法的岩体结构面优势产状聚类分析

宋腾蛟， 陈剑平， 张文， 宋盛渊   

  1. (吉林大学 建设工程学院， 吉林 长春130026)
  • 收稿日期:2013-12-16 修回日期:2013-12-16 出版日期:2015-02-15 发布日期:2014-11-07
  • 通讯作者: 宋腾蛟
  • 作者简介:宋腾蛟(1988-)，男，吉林长春人，吉林大学博士研究生； 陈剑平(1957-)，男，福建南平人，吉林大学教授，博士生导师.
  • 基金资助:
    国家自然科学基金重点资助项目(41330636).

Clustering Analysis of Dominative Attitudes of Rock Mass Structural Plane Based on Firefly Algorithm

SONG Teng-jiao， CHEN Jian-ping， ZHANG Wen， SONG Sheng-yuan   

  1. College of Construction Engineering， Jilin University， Changchun 130026， China.
  • Received:2013-12-16 Revised:2013-12-16 Online:2015-02-15 Published:2014-11-07
  • Contact: CHEN Jian-ping
  • About author:-
  • Supported by:
    -

RichHTML

1

PDF (PC)

938

摘要： 传统的图形分析方法较为粗糙，且主观因素影响较大，因此提出一种基于萤火虫算法的岩体结构面产状数据划分的新方法.用结构面单位法向量表示结构面的产状，通过单位法向量所夹锐角正弦值来衡量结构面间的距离，并建立模糊目标函数，然后运用萤火虫优化算法搜索最优聚类中心，同时确定组间边界.由于萤火虫算法具有出色的全局寻优能力和较快的收敛速度，使得新算法在解决结构面产状划分问题时非常省时高效.最后应用此算法对怒江松塔坝址区平硐实测结构面产状数据进行优势组划分，结果显示，此方法合理有效，优势明显.

关键词: 岩石力学, 结构面, 产状数据划分, 聚类方法, 萤火虫算法

Abstract: Traditional graphic analysis is inadequate and greatly influenced by subjective experience. A new method based on firefly algorithm is proposed to partition the discontinuity orientation data. The sine of the acute angle between the unit normal vector of two discontinuities is used to quantify the distance between the orientation of these discontinuities. Then， the fuzzy objective function is established and firefly algorithm is chosen to search the optimal clustering centers. At the same time， the boundaries between different sets are determined. Since the firefly algorithm has the outstanding ability of global optimization and a fast convergence speed， the new algorithm has the advantages of time saving and high efficiency. Finally， this new method is applied in discontinuity orientation data partition at Songta dam site， Nujiang River. The results indicate that the new approach is reasonable and has obvious superiority.

Key words: rock mechanics, structural plane, attitude data partition, clustering method, firefly algorithm

中图分类号: