摘要
应用基于神经网络的随机有限元,把截面抗压强度条件视为承载能力极限状态,进行某高速公路不等跨连拱隧道衬砌结构可靠度分析,得出了衬砌结构可靠指标的分布规律和衬砌结构的概率破坏规律。进行衬砌结构可靠度对随机变量的敏感性分析,得出随机变量的分布类型与变异系数对可靠指标的影响规律。结论对隧道设计与施工具有参考价值。
A neural network-based model for reliability analysis of unequal dcoble-areh tunnel lining structure was proposed to represent the serviceability performance function. The Monte-Carlo simulation in combination with neural network technique was used to calculate the reliability index of tunnel lining structure. The reliability index in ease of normal distributions is Bigger than that when all the random variables are assumed, ned to be lognormal distributions. Relationship between relisability index and random variable＇s coefficient of variance was also investigated. The results show that the coefficient of variance of elastic modulus exerts the greatest influence on the reliability index of tunnel lining structure.
出处
《地下空间与工程学报》
CSCD
2005年第6期841-843,847,共4页
Chinese Journal of Underground Space and Engineering
基金
国家自然科学基金项目资助(No.50404010
50490274)
湖南省杰出青年科学基金资助
关键词
不等跨连拱隧道
衬砌
可靠度
敏感性分析
unequal double-areh tunnel
lining structures
reliability analysis
sensitivity analysis
