Smart data driven maintenance: Improving damage detection and assessment on aerospace structures

Abstract

Data driven on-line assessment of structural health of aircraft fuselage panels is crucial both in military and civilian settings. This paper shows how Support Vector Machines (SVM) and Genetic Algorithm (GA) enable to analyze the strain values acquired through a monitoring sensor network and improve the diagnostic steps: 1) detecting a damage 2) identifying the specific component affected 3) characterizing the damage in terms of centre and size. The first two steps are performed through the SVM while the 3rd step is based on an Artificial Neural Network (ANN). Finally, the remaining useful life is estimated by using ANNs to predict the values of two parameters of the NASGRO equation which is used to estimate the damage propagation.