STOCHASTIC PARAMETER ESTIMATION OF NON-LINEAR SYSTEMS USING ONLY HIGHER ORDER SPECTRA OF THE MEASURED RESPONSE

✍ Scribed by M. Vasta; J.B. Roberts

• Publisher: Elsevier Science
• Year: 1998
• Tongue: English
• Weight: 361 KB
• Volume: 213
• Category: Article
• ISSN: 0022-460X
• DOI: 10.1006/jsvi.1997.1446

✦ Synopsis

Methods for using fourth order spectral quantities to estimate the unknown parameters in non-linear, randomly excited dynamic systems are developed. Attention is focused on the case where only the response is measurable and the excitation is unmeasurable and known only in terms of a stochastic process model. The approach is illustrated through application to a non-linear oscillator with both non-linear damping and stiffness and with excitation modelled as a stationary Gaussian white noise process. The methods have applications in studies of the response of structures to random environmental loads, such as wind and ocean wave forces.