摘要
为获得一种更有效的估算结构基本自振周期的方法,采用SVSA结构振动信号采集分析系统对上海地区287幢房屋的脉动响应进行了现场实测,得到574个结构脉动响应样本及其对应的结构相关信息;针对现场动测结果数据量多、不规律、变化大的特点,采用基于方差分析的特殊统计处理方法对现场动测数据进行分析回归。经方差分析得到对结构基本自振周期影响最大的3个因素为层数、抗侧力体系与结构材料,以层数为主导因素,分别对其余2个因素分类回归经验公式。在确保回归效果显著的前提下,考虑实际应用的简便性,提出了简化的具有双侧95%置信区间的多因素型周期预测模型。
In order to derive a more efficient method for the estimation of a structural fundamental natural period, 287 buildings in Shanghai are measured by SVSA to obtain 574 ambient vibration response measurements and the structural information are collected as well. Special statistical approach based on ANOVA （analysis of variance） is applied to analyze and to regress the data for its large quantity, irregularity and variability. According to the results of ANOVA, the number of stories, lateral-force-resisting-system and structural material are the three most significant factors to a fundamental natural period, thus empirical formulas are regressed respectively for the latter two while regarding the number of stories as the key factor. On the premise of the significance of regression effects, simplified multi-factor fundamental period predictive models with two-side 95% confidence intervals are proposed for the convenience in application.
出处
《工程力学》
EI
CSCD
北大核心
2013年第3期146-151,共6页
Engineering Mechanics
基金
国家重点基础研究发展计划973项目(2007CB714202)
