Advances in Aerodynamics

Table 7 Comparison of accuracy metrics for predictions of aerodynamic performance by MLP, CNN_2D, and CNN_1D in high-dimensional aerodynamic modeling for flight status and wing shape variables

From: High-dimensional aerodynamic data modeling using a machine learning method based on a convolutional neural network

Object	Model	Relative error ε	R²	RRMSE	RMAE
\({C}_{L}\)	MLP	7.69%	0.9813	0.1366	0.4638
	CNN_2D	3.42%	0.9963	0.0606	0.2998
	CNN_1D	2.62%	0.9978	0.0466	0.1779
\({C}_{D}\)	MLP	9.97%	0.9076	0.3036	3.3977
	CNN_2D	9.58%	0.9147	0.2917	2.3320
	CNN_1D	6.04%	0.9661	0.1840	0.9539
\({C}_{M}\)	MLP	7.16%	0.9815	0.1357	0.5212
	CNN_2D	3.56%	0.9954	0.0676	0.3425
	CNN_1D	2.91%	0.9970	0.0551	0.2391

Back to article page