Open AccessOn the Use of Weighted Least Squares for Time Domain Modal Parameter Identification
Author(s) -
Jonathan Cooper
Publication year - 2004
Publication title -
shock and vibration
Language(s) - English
Resource type - Journals
SCImago Journal Rank - 0.418
H-Index - 45
eISSN - 1875-9203
pISSN - 1070-9622
DOI - 10.1155/2004/142176
Subject(s) - weighting , time domain , impulse response , algorithm , identification (biology) , least squares function approximation , modal , total least squares , range (aeronautics) , computer science , mathematics , system identification , generalized least squares , impulse (physics) , series (stratigraphy) , statistics , engineering , data mining , mathematical analysis , paleontology , estimator , aerospace engineering , chemistry , biology , quantum mechanics , singular value decomposition , computer vision , radiology , measure (data warehouse) , medicine , botany , physics , polymer chemistry
An approach for the use of the Weighted Least Squares method for time domain modal parameter identification is introduced. The approach enables standard identification methods to be employed in the usual manner, except that sections of the time records with high signal-noise ratios are given a greater emphasis in the identification process. It is shown how it is possible to apply the method to a wide range of applications, including the analysis of impulse response, input-output and time varying data. A number of different weighting schemes are explored, based upon exponential weighting and the envelope of the measured data. A series of simulated tests demonstrate how the performance of conventional time domain algorithms can be improved significantly with little extra computational effort though the use of a Weighted Least Squares scheme
Accelerating Research
Robert Robinson Avenue,
Oxford Science Park, Oxford
OX4 4GP, United Kingdom
Address
John Eccles HouseRobert Robinson Avenue,
Oxford Science Park, Oxford
OX4 4GP, United Kingdom