Hamiltonian Neural Network 6-DoF Rigid-Body Dynamic Modeling Based on Energy Variation Estimation

Figure 3

The trajectories of all comparative models and the actual model under identical initial conditions, with aircraft performing the same random action sequences (a). In addition, the mean absolute error (MAE) of all models’ coordinates in relation to the actual model (b) and the energy-time variation curves of all models and the actual aircraft model (c).

(a)

(b)

(c)