欢迎访问《兵工学报》官方网站,今天是

兵工学报 ›› 2007, Vol. 28 ›› Issue (4): 495-498.

• 短文 • 上一篇    下一篇

基于人工神经网络的结构可靠性分析

杨多和,安伟光,李铁钓   

  1. 哈尔滨工程大学航天工程系,黑龙江哈尔滨150001
  • 收稿日期:2005-11-17 上线日期:2014-12-25
  • 通讯作者: 杨多和

Structural Reliability Analysis Based on Artificial Neural Network

YANG Duo-he, AN Wei-guang, LI Tie-jun   

  1. School of Civil Engineering, Harbin Engineering University, Harbin 150001,Heilongjiang, China
  • Received:2005-11-17 Online:2014-12-25
  • Contact: YANG Duo-he

摘要: 在结构可靠性分析中,通过人工神经网络拟合极限状态函数,借助神经网络的函数映射关系产生极限状态函数值。采用Mont。Carlo法随机抽样的思路,对大范围的数据进行概率分析,得到极限状态函数值的均值和标准差,求得结构系统的可靠性指标,该方法为极限状态函数无法显式表达的结构系统可靠性分析问题提供了一个可行方法。

关键词: 人工智能 , 人工神经网络 , 极限状态函数 , 结构可靠性 , Monte Carlo法

Abstract: The limit state function can be fitted by artificial neural network and its value can be gener?ated by function mapping relation of neural netwok for the structure reliability analysis. Adopting the thinking of Monte Carlo random samples, the probability analysis to the data of the big scope can be carried on, the mean value of the limit state function and standard deviation can be got, the reliability index of the structure system can be solved. Under the condition that the limit state function is unable to make explicit expression, a feasiole method was provided for reliability analysis question of the structure system.

Key words: artificial intelligence , artificial neural network , limit state function , structural reliability , Monte Carlo method

中图分类号: