Comparison of optimal sensor placement algorithms via implementation on an innovative timber structure

Abstract

This research work aims at deriving an optimal sensor configuration for the modal identification of an innovative timber structure. The goal is to extract the maximum possible information from the structure, while minimizing the amount of deployed sensors. Several different optimal placement methods are implemented, including the effective independence method (EFI), the modal kinetic energy method (MKE) and the information entropy method (IEI). Additionally, a modified version of the IEI method, incorporating a prediction error correlation term is introduced. This addition alleviates sensor clustering around a single location, which often occurs if the modal input data is generated from a finite element model of a dense mesh of possible sensor positions. The study shows, that to obtain an optimal sensor setup, prediction error correlation effects should be considered in order to avoid sensor clustering, and the multiaxiality of sensors should be taken into account during the optimization process. © 2017 Taylor & Francis Group, London.