On-line environmental noise driven 3-DOF per story parametric identification of a building

摘要

Feasibility of recovering the parameters of a shear building model with three degrees of freedom per story using a least squares with forgetting factor algorithm is analyzed. Two orthogonal horizontal seismic signals, known as environmental noise, are used as excitation for the algorithm. Environmental noise are small magnitude signals induced by several causes: sensor noise, wind effects, machinery operation, vehicular traffic and others phenomena. These signals, in contrast with seismic excitation, do not produce structural damage and are easy to obtain by installing accelerometers on the building. The aim of this work is to show that these signals still contain relevant information related with building dynamic's response, and that with proper use of identification algorithms, precise information about the building model parameters can be obtained for use in the design of vibration control algorithms. As the identification with small magnitude signals can be executed on line before a large earthquake occurs, calculation of proper control signals would not have to wait for the results of an identification excited by a strong seismic signal, saving critical time and helping to avoid structure damage.