We address the problem of structural modal identification duringrnnormal conditions, and this with uncontrolled, unmeasuredrnand nonstationary excitation. We are especially interestedrnin extracting the modal characteristics (frequency, damping,rnmodeshapes) from an in-flight data set. The data can be input/rnoutput or output only. The records can be very short andrnvery noisy. We apply the covariance driven stochastic subspacern(both output-only and input/output) to the flight data extractionrnproblem. We show how important is the analysis ofrnthe result, namely the exploitation of the stabilization diagram,rnespecially in case of highly coupled modes. We also showrnthe implementation and the methodology through a demonstrationrnof the MODAL Scilab toolbox. We show how to obtainrnthe best estimate for the modal parameters, especially for therndampings.
展开▼
