Reinforcement learning for post-coronagraphic wavefront control
Abstract
Direct imaging of exoplanets is limited by the extreme contrast between the star and the planets, which is mitigated using a coronagraph. However, optical aberrations cause starlight leakage through the coronagraph, producing speckles that obscure the planetary signal. Achieving the required contrast levels demands wavefront control with subnanometric precision. Deep reinforcement learning offers a promising alternative to traditional focal-plane wavefront control techniques by enabling adaptive correction strategies learned directly from interaction with the system. In this work, we present a fully data-driven method for post-coronagraphic aberration correction in a simulated high-contrast imaging testbed. The agent controls a deformable mirror using observations consisting of focal-plane measurements (images) and physics-informed wavefront sensing information derived from these images. We evaluate different observation representations and control strategies, and the method is validated on simplified simulations of a high-contrast imaging testbed, where it successfully creates dark holes, i.e., regions of the focal plane in which residual starlight is strongly suppressed, while approaching the performance of conventional wavefront control methods.
Related
- Exploring reinforcement learning to enhance focal-plane wavefront control for vortex coronagraphs
- Focal plane modal estimation of high-order wavefront drifts on coronagraphic testbeds
- Modal estimation of high-order wavefront drifts on coronagraphic testbeds for high-contrast direct imaging of exoplanets
- Pseudo Closed-Loop Bootstrapping for Reinforcement Learning Based AO Control
- Model-based low-order wavefront sensing and control with a fully physical digital twin on MagAO-X
BibTeX
@misc{castanedamedina2026reinforcement,
title = {{Reinforcement learning for post-coronagraphic wavefront control}},
author = {Manuela Castañeda-Medina and Yann Gutierrez and Johan Mazoyer and Baptiste Abeloos and Laurent Mugnier and Olivier Herscovici-Schiller},
howpublished = {arXiv},
year = {2026},
month = sep,
doi = {10.48550/arxiv.2609.20880},
eprint = {2609.20880},
archivePrefix = {arXiv},
url = {https://arxiv.org/abs/2609.20880},
}
RIS
TY - UNPB
TI - Reinforcement learning for post-coronagraphic wavefront control
AU - Manuela Castañeda-Medina
AU - Yann Gutierrez
AU - Johan Mazoyer
AU - Baptiste Abeloos
AU - Laurent Mugnier
AU - Olivier Herscovici-Schiller
PY - 2026
DA - 2026/09/16
PB - arXiv
DO - 10.48550/arxiv.2609.20880
UR - https://arxiv.org/abs/2609.20880
AB - Direct imaging of exoplanets is limited by the extreme contrast between the star and the planets, which is mitigated using a coronagraph. However, optical aberrations cause starlight leakage through the coronagraph, producing speckles that obscure the planetary signal. Achieving the required contrast levels demands wavefront control with subnanometric precision. Deep reinforcement learning offers a promising alternative to traditional focal-plane wavefront control techniques by enabling adaptive correction strategies learned directly from interaction with the system. In this work, we present a fully data-driven method for post-coronagraphic aberration correction in a simulated high-contrast imaging testbed. The agent controls a deformable mirror using observations consisting of focal-plane measurements (images) and physics-informed wavefront sensing information derived from these images. We evaluate different observation representations and control strategies, and the method is validated on simplified simulations of a high-contrast imaging testbed, where it successfully creates dark holes, i.e., regions of the focal plane in which residual starlight is strongly suppressed, while approaching the performance of conventional wavefront control methods.
ER -