Flight Data Analyses of Fiber Optic Based Airworthy Structural Health Monitoring System for UAV using Artificial Neural Networks

摘要

This paper presents an airworthy, Fiber Bragg Gratings (FBG) based, Structural Health monitoring System (SHM) system for an Unmanned Aerial Vehicles (UAV). Various design issues pertaining to sensors location, embedment, integration of interrogation system instrumentation, online data recording, implementation of mathematical models for load estimations and GUI based flight data processing software are addressed. FBG data were processed to identify both vibration modes and loads using signal processing techniques and artificial neural network (ANN) algorithms respectively. The issue of sensor malfunctioning is also addressed wherein sensor failure was incorporated in the in-flight data during post processing for various flight regimes. The ANN based methodology was designed for identification of sensor failure and prediction of the estimated strain based on the available values from working (non-failed) sensors. The performance of load estimation was also compared in both the scenario (i.e. in the event of sensor failure and without sensor failure).