The Innovations Approach to Experimental Modeling of Flexible Structures

Abstract

A technique based on the innovations approach to linear least square identification is proposed for experimental state space modeling of flexible structures. Upon the selection of the model structure and order, the unknown parameters are identified by minimizing the sum of squares of the innovations process. Contrary to having a black-box approach to experimental modeling, this technique uses physical parameterization of the structural dynamics, as well as measurement noise properties. The innovations approach to experimental modeling of the flexible structures automates the identification of unknown parameters in the model and reduces the number of these unknowns to its minimum resulting in small variances in their estimated values. The effectiveness of this experimental modeling technique is demonstrated by accurately identifying the modal information of a rectangular flexible plate, free at two sides and fixed at the other two. The plate is patched with PZT actuators and nearly-collocated sensors at two locations.