C. Yildirim, A. M. Franco Pereira, R. E. Lillo Rodriguez
In this work, a practical reliability analysis and engine health prognostic study is performed using Functional Data Analysis. Multi-sensor data collected from aircraft engines are processed in order to solve one of the most important reliability analysis problems which is estimating the health condition and remaining useful life of an aircraft engine. Smooth sensor curves obtained from different informative sensors are processed using Multivariate Functional Principal Component Analysis to develop a predictive machine learning model and estimate upcoming failures before the failure occurs. Distribution of the principal component scores provided us to understand sensor behavior and suggests a differentiation of different types of machines based on qualitative variables.
Palabras clave: FDA, MFPCA, Predictive Maintenance, Reliability Analysis, Engine Prognostic
Programado
GT06 Análisis de Datos Funcionales I. Nuevas metodologías
8 de junio de 2022 12:40
Salón de Grados
