Eric Hernandez

Applied Observer Theory in Structural Health Monitoring and Diagnostics

Date and time: May 18th 9:00 am
Location: 306 Egan Research Center

ABSTRACT

The objective of the present work is to develop a framework based on the theory of observers to approach two classes of problems typically encountered in monitoring large scale civil engineering structures. The first class consists in reconstructing the dynamic response when model errors constitute the main source of discrepancy between the open loop nominal model estimate and the measurements. To approach this problem we propose the use of a reduced order observer. The properties of the proposed observer are examined and its accuracy compared with other existing methodologies such as Robust Kalman filtering, imposed motions and modal interpolations in the context of instrumented building structures subject to earthquakes. The study considers simulated data and measurements from real structures and in both scenarios linear and nonlinear cases are examined. It is found that the proposed observer outperforms all other methodologies in estimating displacements, velocities, inter-story drifts, shears and overturning moments.

The second category of problems examined is that where uncertainty in the state springs from incomplete information regarding the excitations. To tackle this problem we propose the use of the Kalman filter. As a specific application wind turbine structures are considered. Specifically, we focus on the number of crossings above certain levels of stress at the base of the vertical cantilever that supports the generator. It is shown that the Kalman filter states obtained with minimal levels of instrumentation offer substantial improvements in accuracy when compared with purely stochastic estimates. Noteworthy is the fact that the excitations do not satisfy basic assumptions used in the derivation of either the Kalman filter or the purely stochastic procedures.

Ph.D. Defense Committee:
Prof. D. Bernal (Chairman). CEE-NU
Prof. G. Tadmor. ECE-NU
Prof. B. Shafai. ECE-NU
Prof. L. Caracoglia. CEE-NU
Prof. G. McDaniel. ME-BU