01.1 - Aircraft Design and Integrated Systems (Theory, Methods and Tools)
DEEP LEARNING-BASED INVERSION OF NEAR-FIELD OVERPRESSURE SIGNATURES FOR LOW-BOOM SUPERSONIC TRANSPORT
Y.-H. Peng¹, K.-S. Zhang¹, Q. Chen¹, Z.-H. Han¹; ¹Northwestern Polytechnical University, China
Supersonic transport (SST) aircraft face challenges from sonic booms over land. Low-boom design aims to minimize these effects but is computationally expensive. This study introduces a DNN-based framework to invert ground-level sonic boom signals to near-field overpressure signatures, using autoencoders, PCA, and MLPs. The approach reduces costs, accelerating low-boom SST design.
DOI: 10.71945/icas2026_0278