摘要
针对动力学频率测试中模型的试验值与理论值误差较大的问题,结合大型桥梁缩尺模型试验,采取基于响应面法的有限元模型修正技术,获取了符合结构本真状态的有限元动力修正模型。选取二次多项式作为响应面函数的数学模型,在修正参数的显著性分析和中心复合试验设计基础上,运用最小二乘法对试验设计的样本空间数据进行拟合,以目标达到法为迭代标准,得到了自锚式悬索桥响应面的显示函数关系,并直观地给出典型响应的空间曲面模型。R^2和RMSE指标验证结果表明:经修正后各阶频率峰值误差明显降低;除横向二阶频率外,其余各阶频率修正效果理想,基频最小误差仅为0.96%;基于响应面法的有限元动力模型修正准确可靠;经修正后的有限元模型可作为结构再分析的基准模型。
Due to the reduced scales ratio of model is smaller, there is a larger error between experimental values and theoretical values, employed large bridge reduced scale model test for the first time, adopted the finite element model updating technology base on the response surface method, obtained the finite element dynamic updating model which coincides with the true state of structure. Selected a second degree polynomial as the mathematical model of response surface function, based on the significance analysis of corrected parameters and the central composite experiment design, employed the least squares method to fit the sample space data of experimental design, obtained explicit function expressions for response surface of self-anchored suspension bridge with the iterative standard of achieving target, and intuitively provided the spatial curved surface model of typical responses. R2 and root mean square error model verifying results showed that: the dynamic updating finite element model based on the response surface methodology is accurate and reliable; besides the lateral second-order frequency, the updating results are well, and the measured errors of the fundamental frequency is only 0.96%. the peak value errors of each order frequency are significantly cut down after updating,The updated finite element model can be used as the reference model of the structure further analysis.
出处
《公路交通科技》
CAS
CSCD
北大核心
2014年第8期96-101,共6页
Journal of Highway and Transportation Research and Development
基金
哈尔滨市建设系统科学研究计划项目(201107)
关键词
桥梁工程
自锚式悬索桥
模态试验
响应面法
模型修正
bridge engineering self-anchored suspension bridge modal test response surface method model updating