Experimental test of fault tolerant real time operational modal analysis method by voting algorithm for aircrafts

Abstract

Operational modal analysis (OMA) is a key technique to obtain real-time modal parameters (natural frequency, damping ratio, mode shapes) for aircraft. Due to the nature of an aircraft, an extended flight envelope leads to the flexible structures being highly nonlinear and time-variant. For structural health monitoring and real-time flutter test purposes, automated and robust operational modal analysis methods are required. However, each automated OMA method has its own advantages and disadvantages in specific conditions. In this paper, stochastic subspace identification (SSI), autoregressive poly reference (AR PR), Ibrahim time domain (ITD), eigensystem realization algorithm (ERA) and frequency-spatial domain decomposition method (FSDD) operational modal analysis algorithms were investigated on real-time aircraft flight data. Automated versions of these algorithms are used to obtain real-time modal parameters (natural frequency, damping ratio, and mode shapes). Furthermore, the present work focuses on the development of an enhanced method by utilizing each algorithm simultaneously. The enhanced method aims to calculate the real-time modal parameters without the requirement of users to pick physical modes by voting algorithm between the calculated parameters from each method. Each method and enhanced method were tested for flight data and results were verified by ground vibration test (GVT). It is observed that the voting algorithm improves the robustness of the real-time results in different flight conditions when disadvantageous methods in that specific condition are influenced, by compensating with advantageous methods.