随机荷载比下张弦梁结构基于多线性支持向量机的可靠度分析
发布时间：2018-10-22 06:54
【摘要】：预应力张弦梁结构是大跨度钢结构建筑常见的新型结构之一。因为张弦梁的构件较为简单、自重相对较轻以及外形美观等优点,使得该结构在各种公共建筑当中获得越来越多青睐。因此,为了确保预应力张弦梁结构在不同荷载比值效应下,能够满足规范要求的“能力设计”的规定,特此对预应力张弦梁结构进行分析研究。本文通过ANSYS软件考虑双重非线性对张弦梁结构在不同比例的全跨和半跨荷载作用下的承载性能进行分析,得出了张弦梁结构在不同比值的荷载组合下的承载失效特征。支持向量机是基于统计学习理论而提出的机器学习方法。在可靠度分析方面,支持向量机现已经被应用于与传统可靠度计算方法相结合,并得出了基于支持向量的可靠度算法。本文通过采用多线性支持向量机的策略,即通过以线性方程组替代结构非线性的功能函数,解决以往基于支持向量机的可靠度算法中核函数的选取和参数选优等问题。经过算例对采用基于多线性支持向量机的可靠度算法与采用蒙特卡洛法得到的可靠度指标进行对比,证明基于多线性均匀支持向量机的可靠度算法在可靠度分析中稳定可行性。再将该方法应用于对预应力张弦梁的可靠度分析当中,拟合出张弦梁结构在不同的条件下隐式的线性方程。最后通过引入计算模式不定性的随机参数,在拟合出的线性失效方程的基础上,采用不同的设计方法对预应力张弦梁可靠度进行分析对比。研究表明：随着荷载效应比值的增大,张弦梁结构的可靠指标会明显下降；采用荷载分项系数法对张弦梁结构进行可靠度设计较安全系数校准法更为稳定；当半跨荷载比重较大时,应适当提高张弦梁可靠度设计的安全系数或者荷载分项系数。
[Abstract]:Prestressed beam string structure is one of the common new structures in long span steel structures. Because the beam string structure is relatively simple, the weight is relatively light and the appearance is beautiful, the structure has been more and more popular in various public buildings. Therefore, in order to ensure that the prestressed beam string structure can meet the requirements of "capacity design" of the code under different load ratio effects, this paper makes an analysis and research on the prestressed beam string structure. In this paper, the load-bearing behavior of beam string structure under different ratio of full span and half span loads is analyzed by ANSYS software, and the failure characteristics of beam string structure under different ratio of load combination are obtained. Support vector machine (SVM) is a machine learning method based on statistical learning theory. In the aspect of reliability analysis, support vector machine (SVM) has been applied in combination with the traditional reliability calculation method, and a reliability algorithm based on support vector has been obtained. In this paper, the strategy of multi-linear support vector machine (SVM) is used to solve the problem of kernel function selection and parameter selection in the previous reliability algorithms based on support vector machine by replacing the structural nonlinear functional functions with linear equations. By comparing the reliability index between the multi-linear support vector machine based reliability algorithm and the Monte Carlo method, it is proved that the reliability algorithm based on multi-linear uniform support vector machine is stable and feasible in reliability analysis. Then the method is applied to the reliability analysis of prestressed beam string, and the implicit linear equation of the beam string structure under different conditions is fitted. Finally, the reliability of prestressed beam string is analyzed and compared with different design methods by introducing uncertain random parameters of calculation model and fitting the linear failure equation. The results show that the reliability index of the beam string structure decreases obviously with the increase of the ratio of load effect, and the reliability design of the beam string structure by the method of load subdivision coefficient is more stable than that by the calibration method of safety factor. When the proportion of half span load is large, the safety factor or load component factor of reliability design of beam string should be improved appropriately.
【学位授予单位】：长沙理工大学
【学位级别】：硕士
【学位授予年份】：2014
【分类号】：TU399
本文编号：2286423
[Abstract]:Prestressed beam string structure is one of the common new structures in long span steel structures. Because the beam string structure is relatively simple, the weight is relatively light and the appearance is beautiful, the structure has been more and more popular in various public buildings. Therefore, in order to ensure that the prestressed beam string structure can meet the requirements of "capacity design" of the code under different load ratio effects, this paper makes an analysis and research on the prestressed beam string structure. In this paper, the load-bearing behavior of beam string structure under different ratio of full span and half span loads is analyzed by ANSYS software, and the failure characteristics of beam string structure under different ratio of load combination are obtained. Support vector machine (SVM) is a machine learning method based on statistical learning theory. In the aspect of reliability analysis, support vector machine (SVM) has been applied in combination with the traditional reliability calculation method, and a reliability algorithm based on support vector has been obtained. In this paper, the strategy of multi-linear support vector machine (SVM) is used to solve the problem of kernel function selection and parameter selection in the previous reliability algorithms based on support vector machine by replacing the structural nonlinear functional functions with linear equations. By comparing the reliability index between the multi-linear support vector machine based reliability algorithm and the Monte Carlo method, it is proved that the reliability algorithm based on multi-linear uniform support vector machine is stable and feasible in reliability analysis. Then the method is applied to the reliability analysis of prestressed beam string, and the implicit linear equation of the beam string structure under different conditions is fitted. Finally, the reliability of prestressed beam string is analyzed and compared with different design methods by introducing uncertain random parameters of calculation model and fitting the linear failure equation. The results show that the reliability index of the beam string structure decreases obviously with the increase of the ratio of load effect, and the reliability design of the beam string structure by the method of load subdivision coefficient is more stable than that by the calibration method of safety factor. When the proportion of half span load is large, the safety factor or load component factor of reliability design of beam string should be improved appropriately.
【学位授予单位】：长沙理工大学
【学位级别】：硕士
【学位授予年份】：2014
【分类号】：TU399
【参考文献】
相关期刊论文 前1条
1 刘锡良,白正仙;张弦梁结构的有限元分析[J];空间结构;1998年04期
，本文编号：2286423
本文链接：https://www.wllwen.com/guanlilunwen/chengjian/2286423.html
