An adaptive data sampling strategy for stabilizing dynamical systems via controller inference
This work addresses the problem of data collection for stabilizing controllers in engineering applications, particularly in edge cases and limit states where instabilities make data collection difficult, representing a domain-specific incremental improvement.
The paper tackles the challenge of collecting informative data for learning stabilizing controllers in unstable dynamical systems by proposing an adaptive sampling scheme that simultaneously stabilizes the system during data collection. The result is a method that learns controllers with up to one order of magnitude fewer data samples than unguided approaches.
Learning stabilizing controllers from data is an important task in engineering applications; however, collecting informative data is challenging because unstable systems often lead to rapidly growing or erratic trajectories. In this work, we propose an adaptive sampling scheme that generates data while simultaneously stabilizing the system to avoid instabilities during the data collection. Under mild assumptions, the approach provably generates data sets that are informative for stabilization and have minimal size. The numerical experiments demonstrate that controller inference with the novel adaptive sampling approach learns controllers with up to one order of magnitude fewer data samples than unguided data generation. The results show that the proposed approach opens the door to stabilizing systems in edge cases and limit states where instabilities often occur and data collection is inherently difficult.