EUVIMEDEUVIMED entdecken
Crossref

Multi-objective Bayesian optimization approach for hyperparameter tuning in wavefront-sensorless adaptive optics

Zhiguang Zhang, Chang Liu, Zhaojun Yan, Ze Wang, Huizhen Yang

Applied Optics · 20.06.2026 · 65 (18) · 6114

DOI: 10.1364/ao.588137 ↗

Kurz gefasst

Fachartikel in der Zeitschrift Applied Optics, online erschienen am 11.06.2026. Verfasst von Zhiguang Zhang und 4 weiteren Autor:innen. Die Arbeit nennt 26 Literaturangaben. Gefördert von: National Key Laboratory of Adaptive Optics Foundation, National Natural Science Foundation of China.

AbstractBibliografische AngabenZitierenVerwandte Studien

Abstract (Originalsprache)

Wavefront sensorless (WFSless) adaptive optics (AO) systems have been widely used in laser communication and astronomical imaging. Their performance strongly depends on algorithmic hyperparameters, and optimizing these parameters requires balancing multiple performance metrics, generally including the Strehl ratio (SR), power in the bucket (PIB), full-width at half-maximum (FWHM), and convergence speed. In this paper, we propose a multi-objective Bayesian optimization (MOBO) framework for hyperparameter tuning in WFSless AO. A Gaussian process model for multiple objectives is established to capture performance trade-offs, while a logarithmic variant of the Expected Hypervolume Improvement acquisition function is employed to efficiently and stably update the Pareto front. Simulation and experimental results demonstrate that the proposed approach significantly reduces hyperparameter tuning time, achieving comparable or superior performance with more than an order-of-magnitude fewer tuning evaluations than grid search. This work provides both theoretical insight and practical guidance for applying Bayesian optimization to WFSless AO control systems.

Förderung

  • National Key Laboratory of Adaptive Optics Foundation
  • National Natural Science Foundation of China

Zitieren

APA

Zhang, Z., Liu, C., Yan, Z., Wang, Z., & Yang, H. (2026). Multi-objective Bayesian optimization approach for hyperparameter tuning in wavefront-sensorless adaptive optics. Applied Optics, 65(18), 6114. https://doi.org/10.1364/ao.588137

Vancouver

Zhang Z, Liu C, Yan Z, Wang Z, Yang H. Multi-objective Bayesian optimization approach for hyperparameter tuning in wavefront-sensorless adaptive optics. Applied Optics. 2026;65(18):6114. doi:10.1364/ao.588137

BibTeX
@article{Zhang2026,
  author = {Zhang, Zhiguang and Liu, Chang and Yan, Zhaojun and Wang, Ze and Yang, Huizhen},
  title = {Multi-objective Bayesian optimization approach for hyperparameter tuning in wavefront-sensorless adaptive optics},
  journal = {Applied Optics},
  year = {2026},
  volume = {65},
  number = {18},
  pages = {6114},
  doi = {10.1364/ao.588137}
}
Kostenlos starten →