输入信息不完备情况下的混和修正识别法  

STUDY ON MIXED CORRECTION IDENTIFICATION METHOD WITH INCOMPLETE INPUT INFORMATION

作　　者：王建有[1] 黄景忠[1] 陈健云[2] 刘金云[2] 

机构地区：[1]郑州大学环境与水利学院,郑州450001 [2]大连理工大学土木水利学院,大连116023

出　　处：《振动与冲击》2007年第5期123-125,129,共4页Journal of Vibration and Shock

摘　　要：回顾输入信息测试不完备条件下的结构识别问题,并对近年来提出的基于最小二乘类的时域复合反演算法做了深入研究,在统计平均算法的基础上提出了加权平均算法。同时针对满足多个修正条件的结构情况,结合补偿算法,提出混和修正识别法。最后以地震动未知情况下的桁架桥结构为例来说明了算法的可行性。Structural identification with incomplete input information is reviewed,and the multiple inversion method based on the least square method is further studied.Based on the statistical average approach,a new weighted average approach for the multiple inversion method is put forwarded.Further more,for structures with more than one correction condition,combined with the compensation approach,a new approach called mixed correction identification method is proposed.At last,as numerical example,a truss bridge with unknown ground motion is identified and the results show the effectiveness of the proposed method.

关 键 词：系统识别 结构识别 复合反演 信息不完备 

分 类 号：TU311.3[建筑科学—结构工程]