Event
Disturbance-aware Koopman model predictive control framework proposed for quadrotors
1 reports · 1 sources
Stella Kombo et al. proposed the Disturbance-Aware Koopman Control Family (DA-KCF) framework for trajectory tracking control of quadrotors under spatially varying wind fields: it retains an offline-identified nominal model, learns only disturbance corrections online, and embeds model predictive control (MPC) to optimize position and yaw references. The reference-dependent KCF in this framework is first learned using simulation data and then calibrated with limited hardware data; at deployment, the nominal model is fixed, and a lightweight residual ensemble learns low-dimensional disturbance corrections from causal state-command-state transitions. Hardware experiments on a micro quadrotor (uniform wind and spatially varying wake) showed that as disturbance-induced model mismatch increased, online adaptation improved multi-step prediction and trajectory tracking, while meeting real-time control requirements.
Report timeline
- Real-Time Disturbance-Aware Koopman Model Predictive Control for Quadrotors in Spatially Varying Winds
The study proposes a disturbance-aware Koopman control family (DA-KCF) framework that preserves an offline-identified nominal model, learns only a disturbance correction online, and embeds it into model predictive control (MPC) to optimize position and yaw references. The reference-dependent KCF is first learned from simulation data and then calibrated with limited hardware data; during deployment the nominal model remains fixed, and a lightweight residual ensemble learns a low-dimensional disturbance correction from causal state-command-state transitions.
arXiv cs.RO ·