Model parameter identification using an unknown coloured random input

✍ Scribed by J.E. Cooper; M.J. Desforges; J.R. Wright

• Book ID: 102615334
• Publisher: Elsevier Science
• Year: 1995
• Tongue: English
• Weight: 318 KB
• Volume: 9
• Category: Article
• ISSN: 0888-3270
• DOI: 10.1006/mssp.1995.0051

✦ Synopsis

In cases where the input to a system is unknown or unmeasured, assumptions have to be made as to the nature of the excitation in order that modal parameters can be estimated from the response data alone. Commonly, an input signal is assumed to be 'white' random, as is any corruption of the response data due to measurement noise. This work will consider the effects of non-white, 'coloured' input signals upon modal parameter estimates. The correlation fit method will be used to estimate accurate system frequencies and damping ratios from a knowledge of only the response to coloured inputs, including stable-pole signals and force drop-out. Both white and coloured measurement data corruption are considered, and the improved approach demonstrated on simulated data sets.