摘要
针对最小二乘复频域法在噪声干扰下模态参数识别精度不高的问题,采用分母替换函数,提出有理函数拟合改进算法,将其迭代后引入最小二乘复频域法.通过简支梁实验台锤击信号得到频域响应函数,代入到改进的算法中进行研究.结果表明:有理函数拟合最小二乘复频域法比最小二乘复频域法在噪声干扰下的识别效果较好,经过有理函数拟合改进后的多项式在阶次较低的情况下能识别出全部极点并得到更理想的稳态图,且对阻尼比的识别更接近真实值.
The least squares complex frequency domain method is inaccurate in the system modal pa- rameter identification when noise signal exists. Therefore denominator replacement function was proposed. Then rational function fitting iteration was introduced into the least squares complex frequency domain method. Frequency response functions were obtained on a simply supported beam test rig by hammering method. Next they were used into the improved algorithm. It can be drawn that the ra- tional function fitting complex frequency domain method can obtain all poles and much more steady state map on low order polynomial calculation than that of least squares complex frequency domain method under noise signal interference. Besides, the improved method also does well in damping ratio identifying.
出处
《华中科技大学学报(自然科学版)》
EI
CAS
CSCD
北大核心
2014年第5期30-33,46,共5页
Journal of Huazhong University of Science and Technology(Natural Science Edition)
基金
国家高技术研究发展计划资助项目(2008AA042802)
高性能复杂制造国家重点实验室自主研究课题(zzyjkt2013-0613)
关键词
最小二乘复频域法
有理函数
拟合
模态参数
简支梁
least squares complex frequency domain method
rational function
fitting
modal param-eter
simply supported beam