基于支持向量机和子集模拟法的地铁隧道可靠度

发布时间：2018-07-21 21:00

【摘要】：在城区中修建地铁隧道对安全性有极高的要求,传统的安全系数法虽然能在一定程度上反映工程的安全性,但无法考虑变量的随机性。针对传统子集模拟法的不足,引入了一种改进方法,对功能函数阀值处存在相等情况时的子集模拟法,提供一种解决思路。为解决子集模拟法在初始样本点少时的计算不准确问题,基于支持向量机的小样本学习能力,提出随机有限元法(SFEM)、支持向量机(SVM)和子集模拟法耦合的可靠度分析方法。使用耦合方法和传统方法分别计算衬砌单元的失效概率。并用经典算例与传统方法进行了对比,验证了其准确性和应用的可行性,可以作为样本点不足时分析地铁隧道的可靠度方法。结果表明,在样本数量较少的条件下,耦合算法具有很好的计算效率和精度,适用于隧道工程可靠度分析。
[Abstract]:The construction of subway tunnel in urban area has very high requirements for safety. Although the traditional safety factor method can reflect the safety of the project to a certain extent, it can not consider the randomness of variables. In view of the shortcomings of the traditional subset simulation method, an improved method is introduced, which provides a solution to the subset simulation method when the threshold of the function is equal. In order to solve the problem of inaccuracy of the subset simulation method when the initial sample points are small, based on the small sample learning ability of support vector machine, a reliability analysis method based on the coupling of stochastic finite element method (SFEM), support vector machine (SVM) and subset simulation method is proposed. The failure probability of lining element is calculated by coupling method and traditional method. The accuracy and feasibility of its application are verified by comparing the classical example with the traditional method. It can be used as a reliability analysis method for subway tunnel when the sample points are insufficient. The results show that the coupling algorithm has good computational efficiency and accuracy under the condition of small number of samples and is suitable for the reliability analysis of tunnel engineering.
【作者单位】：南阳理工学院土木工程学院;
【分类号】：U451;U231

【相似文献】