01.1 - Aircraft Design and Integrated System (Basics and Theory)
AIRCRAFT DESIGN PARAMETER ESTIMATION USING DATA-DRIVEN MACHINE LEARNING MODELS
S. Shinš, S. Leeš, C. Son, Cheongju University, South Korea; K. Yee, Seoul National University, South Korea; šKorea Institute of Industrial Technology, South Korea
This research presented the validity of applying a data-driven approach to the aircraft initial sizing problem. The limitations of the existing methods are overcome by using machine learning techniques. K-nearest neighbors algorithm and Variational autoencoder were used in this study for the imputation of the incomplete aircraft data. Both of the models have successfully estimated target values.