Load and deflection estimation of a fast catamaran towing-tank model via reduced order modeling and optimal natural observer

This paper presents a reliable methodology that allows using large (in time) but sparse (in space) data from sensors to get a full picture of the dynamic response of a scaled model catamaran structure. Increasing the amount of information relative to the structural behavior is critical for some applications like structural health monitoring and condition-based maintenance. Loads and elastic deflections are estimated by combining the information obtained from the measured structural data with a finite element model of the system given that some statistical properties of the estimated outputs are preliminary known. Craig-Bampton method is used to achieve a reduced-order model since it is particularly suited to describe the modal response along with the estimation of the concentrated forces. A natural second-order observer is then defined to provide an optimal estimation of forces and elastic deflections. The validation of this technique is pursued by comparing estimated and known values of the response and a sensitivity analysis of the accuracy is considered as well.