摘要
为了研究高维随机参数作用下大跨度悬索桥运营阶段挠度可靠度,在有限元计算基础上,基于支持向量机(support vector machine,SVM)建立成桥阶段挠度可靠度模型,结合优化后的粒子群算法(Particle Swarm Optimization,PSO)计算结构运营阶段挠度可靠指标。研究结果表明:借鉴遗传算法中的变异思想,通过设置中间变量约束条件,可以解决粒子群算法易早熟、后期迭代效率低的问题,进而提高计算效率与精度,基于SVM-PSO算法的结构可靠度方法高效准确,普立特大桥挠度可靠度满足正常使用极限条件下的要求。
In order to study the reliability of deflection for long-span suspension bridge at operation stage with high dimensional random parameters,this paper established bridge deflection reliability model based on SVM(support vector machine)as well as using finite element method.Combined with the PSO(particle swarm optimization),the reliability index of deflection in structural operation stage was calculated.The result shows that by using the idea of mutation in the genetic algorithm,the problem of premature convergence and low efficiency of the particle swarm optimization can be solved by setting the constraints of intermediate variables,thus improving the efficiency and accuracy of the algorithm.Structural reliability method based on SVM-PSO algorithm is efficient and accurate,and the deflection reliability of the bridge meets the requirements under the normal use limit condition.
作者
邓海波
常柱刚
李胡涛
张红显
DENG Haibo;CHANG Zhugang;LI Hutao;ZHANG Hongxian(Changsha Planning & Design Institute Co.,Ltd,Changsha 410007,China;T.Y.Lin International Engineering Consulting (China) Co.,Ltd,Chongqing 404100,China)
出处
《铁道科学与工程学报》
CAS
CSCD
北大核心
2019年第1期114-120,共7页
Journal of Railway Science and Engineering
基金
国家自然科学基金资助项目(51678072
51478472)
关键词
支持向量机
粒子群算法
可靠度
挠度
有限元法
support vector machine
particle swarm optimization
reliability
deflection
finite element method
