Event
Dual-Gradient Consensus Improves Adaptive Control of Quadrotors
1 reports · 1 sources
Mingtian Tan and Vasileios Tzoumas propose a dual-gradient consensus (DGC) adaptive learning method for quadrotor trajectory tracking under mixed in-domain and short-term out-of-domain disturbances. The method retains DAPL's random Fourier feature model, offline BPTT strategy, and least-squares adaptation, changing only the online update step size: it uses long- and short-term gradient memory to discriminate adaptation states and employs a confidence map to suppress unreliable updates. In numerical quadrotor trajectory tracking, the authors test with residuals from four mixed in-domain components and short-term out-of-domain events. Compared with L1-augmented MPC, RFF-based MPC, fixed-step-size DAPL, and a frozen-parameter baseline, DGC achieves the lowest average position RMSE in all four cases. These results come from the authors' numerical experiments.
Report timeline
- Dual Gradient Consensus (DGC): An Adaptive Learning Method for Transient Out-of-Distribution Disturbances in Quadrotor Control
To address the problem that short-term out-of-distribution residuals can corrupt the learned disturbance model during quadrotor trajectory tracking, DGC improves the online update step size while retaining the random Fourier feature model, the offline BPTT policy, and least-squares adaptation (DAPL). It distinguishes adaptation states by comparing long- and short-term gradient memories and uses a confidence map to suppress unreliable updates.
arXiv cs.RO ·