Table 3 Average absolute initial trim offsets for training set (\(N=252\)) with 95% CI and p-value
From: A physics-based neural network for flight dynamics modelling and simulation
FlyNet | Classical | p - value | |
|---|---|---|---|
\(\dot{u}_0\) [ft\(s^{-2}\)] | 0.861 [0.122, 4.720] | 2.159 [0.164, 5.627] | 0.000 |
\(\dot{v}_0\) [ft\(s^{-2}\)] | 0.824 [0.121, 4.179] | 1.010 [0.110, 4.148] | 0.001 |
\(\dot{w}_0\) [ft\(s^{-2}\)] | 1.259 [0.157, 6.071] | 2.780 [0.230, 8.550] | 0.000 |
\(\dot{p}_0\) [rad\(s^{-2}\)] | 0.027 [0.002, 0.080] | 0.165 [0.014, 0.551] | 0.000 |
\(\dot{q}_0\) [rad\(s^{-2}\)] | 0.010 [0.001, 0.033] | 0.080 [0.009, 0.314] | 0.000 |
\(\dot{r}_0\) [rad\(s^{-2}\)] | 0.015 [0.001, 0.047] | 0.048 [0.004, 0.132] | 0.000 |