中文核心期刊
CSCD来源期刊
中国科技核心期刊
RCCSE中国核心学术期刊

重庆交通大学学报(自然科学版) ›› 2010, Vol. 29 ›› Issue (1): 98-102.

• • 上一篇    下一篇

泥石流危险性SIGA-BP神经网络评价方法及应用

谷秀芝,陈洪凯,刘厚成   

  1. 重庆交通大学岩土工程研究所,重庆 400074
  • 收稿日期:2009-06-30 出版日期:2010-02-15 发布日期:2015-01-22
  • 作者简介:谷秀芝(1982—),女,河北邢台人,硕士研究生,主要从事地质灾害与防治研究。E-mail:gxzh1982@126.com。
  • 基金资助:
    国家自然科学基金项目(50678182);中国博士后科学基金项目(20080430095)

Method and Application of Debris Flow Hazard Assessment Based on SIGA-BP Neural Network

GU Xiu-zhi,CHEN Hong-kai,LIU Hou-cheng   

  1. Institute of Geotechnical Engineering,Chongqing Jiaotong University,Chongqing 400074,China
  • Received:2009-06-30 Online:2010-02-15 Published:2015-01-22

摘要: 泥石流危险度是由泥石流危险因子综合判定的,然而危险因子有主次之分,要从众多泥石流危险因子中筛选 出作用最大的主要危险因子是很困难的,利用自适应免疫遗传算法SIGA(Self-adaptiveImmuneGeneticAlgorithm) 对BP神经网络进行优化,获得了与云南省最相关的7项泥石流危险因子,建立了基于SIGA的BP神经网络模型, 并对10组泥石流沟数据进行预测,得到了较高的预测结果。

关键词: 泥石流, 神经网络, 危险度, 危险因子

Abstract: The risk degree of debris flow is determined by dangerous factors of the debris flow. The dangerous factors are divided into primary and secondary factors. It is difficult to choose the most dangerous factor. BP neural network is optimalized by self-adaptive immune genetic algorithm (SIGA),and seven dangerous factors of Yunnan province are obtained. SIGA-BP neural network is also established,which is applied to forecasting data of 10 groups’debris flow,and more accurate forecasting results are obtained.

Key words: debris flow, neural network, risk degree, risk factors

中图分类号: