Identification of elastic boundary conditions of light-weight device casing walls using experimental data

Device and machinery noise can be reduced by controlling vibration of casing walls - such approach, referred to as the active casing approach, has been successfully applied by the authors in previous publications. It was observed that for an effective active control it is crucial to mount sensors and actuators at appropriate locations on the vibrating structure. However, an optimization process of sensors and actuators arrangement requires a precise mathematical model of such structure. In this research a cuboid light-weight casing is considered. A mathematical model describing walls of such casing requires parameters to be defined that cannot be delivered by the producer nor measured directly. The aim of this paper is to develop and evaluate an identification procedure of model parameters basing on experimental data - distinguished natural frequencies and modeshapes. A memetic algorithm is used to find optimal set of model parameters. Automatic classification of simulated modeshapes is employed. Obtained results are validated experimentally for multiple casing walls.