Towards intelligent fiber laser design by using a feed-forward neural network

Xinyang Liu, Regina Gumenyuk

Tutkimustuotos: KonferenssiartikkeliTieteellinenvertaisarvioitu

9 Lataukset (Pure)

Abstrakti

We demonstrated a high accuracy prediction of the fiber laser output parameters by using a feed-forward neural network. We explored both the gain and spectral filter parameters to test the prediction performance of the neural network and realized the mapping between cavity parameters and laser output performance. We also investigated how the number of hidden layers could influence the accuracy of prediction. Based on the results, the output spectrum and temporal pulse profiles can be predicted with high accuracy in various fiber laser designs. Our work paves the way to intelligent laser design with ultimate autonomy.

AlkuperäiskieliEnglanti
OtsikkoAdvanced Lasers, High-Power Lasers, and Applications XIV
ToimittajatJun Liu, Shibin Jiang, Ingmar Hartl
KustantajaSPIE
ISBN (elektroninen)9781510667693
DOI - pysyväislinkit
TilaJulkaistu - 2023
OKM-julkaisutyyppiA4 Artikkeli konferenssijulkaisussa
TapahtumaSPIE/COS Photonics Asia technical conferences - Beijing, Kiina
Kesto: 14 lokak. 202316 lokak. 2023

Julkaisusarja

NimiProceedings of SPIE - The International Society for Optical Engineering
Vuosikerta12760
ISSN (painettu)0277-786X
ISSN (elektroninen)1996-756X

Conference

ConferenceSPIE/COS Photonics Asia technical conferences
Maa/AlueKiina
KaupunkiBeijing
Ajanjakso14/10/2316/10/23

Julkaisufoorumi-taso

  • Jufo-taso 0

!!ASJC Scopus subject areas

  • Electronic, Optical and Magnetic Materials
  • Condensed Matter Physics
  • Computer Science Applications
  • Applied Mathematics
  • Electrical and Electronic Engineering

Sormenjälki

Sukella tutkimusaiheisiin 'Towards intelligent fiber laser design by using a feed-forward neural network'. Ne muodostavat yhdessä ainutlaatuisen sormenjäljen.

Siteeraa tätä