Identification of Statistical Energy Analysis Parameters from Measured Data

Abstract

An approach for identifying statistical energy analysis (SEA) parameters from experimental investigation is presented. Specifically, a power flow realization method (PRM) and statistical energy analysis model improvement (SMI) technique using transient time-domain vibration measurements are derived. The efforts are refined and validated using a range of test simulations, and then with true physical tests conducted on both simple and complex structures. Experimentation is also used to define the necessary input power measurements, response energy measurements, and data processing techniques necessary for successful PRM/SMI.

It is found that utilization of time domain data allows for an over-determined power balance providing favorable numerical conditions for the identification. In fact, it is observed that a full matrix of measured inputs and outputs is not necessarily required for successful identification as is the case with current methods. Additionally, useful insight into system dimensionality is obtained during the identification process. Furthermore, it is found that the procedure indicates true parameters that are easily distinguished from those associated with noise in the data and, hence, is well suited for this application. Results indicate that the methodology has the potential to significantly enhance standard SEA procedures.