location: Current position: Home >> Scientific Research >> Paper Publications

基于互相关函数幅值和SVM的输电塔损伤识别

Hits:

Indexed by:期刊论文

Date of Publication:2017-06-15

Journal:振动、测试与诊断

Included Journals:Scopus、EI、CSCD

Volume:37

Issue:3

Page Number:495-500

ISSN No.:1004-6801

Key Words:互相关函数;支持向量机;损伤识别;输电塔

Abstract:针对目前输电塔结构损伤识别中需要布设大量传感器的问题,提出了基于互相关函数幅值和支持向量机(support vector machine,简称SVM)的损伤识别方法.首先,定义初始与当前状态结构模态响应近似信号的互相关函数幅值差为损伤特征;其次,将损伤特征作为输入样本来训练支持向量机分类器,将损伤识别问题转化为模式分类问题;最后,利用2层角钢塔模型的振动试验,验证了方法的可行性.该方法仅需要少量传感器测得结构的动力响应,且适用于环境荷载激励,对输电塔结构损伤有较好的识别效果和噪声鲁棒性.

Pre One:Optimum Value of Negative Stiffness and Additional Damping in Civil Structures

Next One:Impedance based bolt pre-load monitoring using piezoceramic smart washer