Conditional Invertible Neural Networks for Data-Driven UAV Control: A 2-D Proof of Concept
For multirotor control, this work demonstrates a proof-of-concept for cINNs as probabilistic inverse-dynamics models, but the results are incremental with limited generalization.
Conditional invertible neural networks (cINNs) were used as probabilistic inverse-dynamics models for multirotor control, achieving open-loop R²=0.944 and closed-loop position RMSE comparable to INDI (9.7 vs. 9.5 m) in 15 scenarios, with 47% tracking acceptable.
We investigate conditional invertible neural networks (cINNs) as probabilistic inverse-dynamics models for multirotor control. For a planar X8 coaxial multicopter, we learn $p(u \mid s_t, c_t)$ from an incremental nonlinear dynamic inversion (INDI) teacher using rational-quadratic spline coupling and invertible linear mixing. Open-loop reproduction reaches $R^2 = 0.944$, mean CRPS 0.0915, and log-probability-error correlation $ρ= -0.60$. Over 15 closed-loop scenarios, position RMSE matches INDI (9.7 vs. 9.5 m), with 47 percent tracking acceptably; failures separate into attitude divergence under aggressive steps and phase lag under high-frequency references, isolating command bandwidth and data coverage as dominant failure mechanisms.