Research in Robust Frequency ResponseFunction Estimator
碩士 === 逢甲大學 === 自動控制工程所 === 90 === The purpose of this paper is to find a robust estimator for getting more precisely frequency response function (FRF) in the system. The traditional FRF estimators treat the system disturbance as Gaussian white noise rather than colored noise in reality. It causes t...
| Main Authors: | , |
|---|---|
| Other Authors: | |
| Format: | Others |
| Language: | zh-TW |
| Published: |
2002
|
| Online Access: | http://ndltd.ncl.edu.tw/handle/yuxaua |
| Summary: | 碩士 === 逢甲大學 === 自動控制工程所 === 90 === The purpose of this paper is to find a robust estimator for getting more
precisely frequency response function (FRF) in the system. The traditional FRF
estimators treat the system disturbance as Gaussian white noise rather than colored
noise in reality. It causes the traditional FRF estimators being not efficiency for
noise suppression. Therefore this research is to cite the concept of H_inf norm in
robust control and using Nearly NP (Nevanlimma-Pick) Interpolation Algorithm to identify the real FRF.
For comparison the robustness of the traditional FRF algorithm (Hv) and
Nearly NP Interpolation Algorithm (Hnp), numerical simulation chooses a highly
damped and highly coupled lumped system contaminated with the Gaussian white
noise or colored noise as a test bench. Further, using this estimated frequency
response functions evaluate system’s natural frequency and damping ratio by CMIF
(Complex Mode Indication Function) algorithm. The results of system’s natural
frequency and damping ratio show that Nearly NP Interpolation Algorithm is more
robust than the traditional FRF estimator is.
|
|---|