Adaptive on-line System Identification of Aerospace Structures using MX Filters.

Abstract

This paper describes a new method of system identification of dynamic structures, which is based on an adptive signal processing that considers structural dynamic aspects and can thus be realized as an on-line procedure. The principle, performance, and realization of the new technique are shown. System identification experiments on a "smart" structure, a carbon fiber reinforced epoxy beam with embedded NiTi wires, demonstrates that the structural parameters to be identified are known each time, even in cases where they are time-varying. Further applications of the new tool are given in the areas of structural health monitoring and controller observations.