基于分布式传感网络的结构模态识别方法.pdfVIP

43 3 Vol. 43 No. 3 2 0 1 0 3 CHINA C IV IL ENGINEER ING JOURNAL M ar. 2010 1 2 3 3 张 敏 谢慧才 Sung han Smi B F S encer Jr ( 1. , 515063; 2. , 518063; 3. , 61801, ) : , , , , , , ( MAC) , : ; ; ; ; : TU311 : A : 1000131X ( 2010) M odal Identification Technique Based on D ist ibuted Senso N etwo ks 1 2 3 3 Zhang M in X ieH u icai SungH an Sim B F Sp encer Jr ( 1. Shantou Un iversity, Shantou 515063, China; 2. Shenzhen U niversity, Shenzhen 518060, China; 3. Un iversity of Illinois atU rbanaCham aign, IL 61801, USA ) Abst act: Identification of the dynam ic characteristics of civil structures is mi ortant in structural health mon itoring. A large number of data must be available from a dense array of sensors for largescale civil structures and oses a b ig challenge to the conventional centralized rocessing technique. Smart sensor netw orks ( SSN ) w ith decentralized rocessing ca ab ility rovides new ossib ilities for structural health mon itoring. A distributed method is ro osed to calcu late the globalmode sha e in SSN. Stochastic subs ace identification is mi lemented to identify localmode sha es, w hich are rescaled by using article sw arm o tmi ization method, and subsequently to combine globalmode sha es. U sing an arch bridge model as an exam le, the distributed method is show n to be effective. The globalmode sha es are close to those from centralizedmethod according to modal assurance criterion (MAC). K eywo ds: structural health mon itoring; smart sensor newt ork; modal identification; centralized technique; distributed m