An Introduction to the "Emphasized Data Analysis" Using the Eigensystem Realization Algorithm.

Abstract

In this report a supplementary technique to improve modal results of overview identification is presented. This new technique, called the Emphasized Data Analysis, combines the basic idea of the traditional phase resonance method with the fast data acquisition andanalyses of multiple input-multiple output techniques (MIMO) in the time domain. To apply the Emphasized Data Analysis, a priori knowledge about the test specimen is required, which usually is provided by a standard MIMO identification, taking into account all available inputs and outputs. While low global accuracy indicators highlight candidate modes for improvement, local accuracy indicators are used for selection of optimal inputs and outputs. The successful application of the Emphasized Data Analysis to the CLUMOD structure is illustrated in this report. Accuracy and stability of modal parameters of six target modes could considerably be improved. In addition, the computation time of the Emphasized Data Analysis is much shorter compared to the standard MIMO approach, due to the fact that only a subset of all available inputs and outputs is used. The Emphasized Data Analysis is very promising, especially for modal testing with large numbers of inputs and outputs.