Non‐parametric identification of a class of non‐linear close‐coupled dynamic systems

Abstract

A non‐parametric identification technique for the identification of arbitrary memoryless non‐linearities has been presented for a class of close‐coupled dynamic systems which are commonly met with in mechanical and structural engineering. The method is essentially a regression technique and expresses the non‐linearities as series expansions in terms of orthogonal functions. Whereas no limitation on the type of test signals is imposed, the method requires the monitoring of the response of each of the masses in the system. The computational efficiency of the method, its easy implementation on analogue and digital machines and its relative insensitivity to measurement noise make it an attractive approach to the non‐parametric identification problem.

An identification methodology for a class of hysteretic structures

“…where the right-hand side of equation (16) is known. The identification problem may, therefore, be re-posed as the identification of the generalized restoring force h' .…”

Obtaining refined first‐order predictive models of linear structural systems

“…Instead of updating an initial model, it is also possible to identify the second-order structural matrices (mass, damping, and sti ness) directly as discussed, among others, by Agbabian et al [4]; however, this methodology, as all similar equation-error approaches that are limited by the number of available sensors and=or that require numerical integration (and=or di erentiation) of available time histories, often leads to poor predictive models (Smyth and Pei [5] provide a lucid analysis of the e ects of using integrated noisy data for various system identiÿcation approaches). Some other noteworthy e orts on system identiÿcation in the civil engineering literature are the works by Ghanem and Shinozuka [6], Hoshiya and Sutoh [7], Masri et al [8], S afak [9], Udwadia and Kuo [10], Lin et al [11], Koh and See [12], and Beck and Katafygiotis [13].…”