Exploiting Over-Approximation Errors as Preview Information for Nonlinear Control*
Antoine Aspeel, Antoine Girard, Thiago Alves Lima · 2025
Read Exploiting Over-Approximation Errors as Preview Information for Nonlinear Control* by Antoine Aspeel, Antoine Girard, Thiago Alves Lima (2025) in the Tower of Babel, a free online library of public-domain works preserved in clean, searchable text. No sign-up. The full text is served as a plain-text file at a predictable URL, where text is present: https://antarctic-labs.com/tower-of-babel/arxiv-2511-03577.txt.
Description
Antoine Aspeel, Antoine Girard, and Thiago Alves Lima turn approximation error from a nuisance into a preview signal for nonlinear control. Controllers often steer a simpler over-approximation of a hard system, then treat the gap as an unknown disturbance. The authors note the gap depends on the input itself, so it carries usable information about what comes next.
Policies that read both state and that error are called informed policies. Turning such a policy back into a valid input for the true system becomes a fixed-point problem. Existence follows from a classic theorem, and efficient solvers cover affine and general nonlinear cases. Using error as preview could tighten control of constrained systems like robots and power converters.
Catalog record
| CREATOR | Antoine Aspeel, Antoine Girard, Thiago Alves Lima |
|---|---|
| YEAR | 2025 |
| COLLECTION | PAPERS |
| FORMAT | TXT |
| FILE SIZE | 32 KB |
| VERSION | 1.0 |
| LICENSE | http://creativecommons.org/licenses/by/4.0/ |
| RIGHTS | CC BY |
| SOURCE | arXiv |
Source: https://arxiv.org/abs/2511-03577