[9a-F211-1~12] 18.2 Process Informatics

11:15 AM - 11:30 AM JST(2:15 AM - 2:30 AM UTC)Presentation by Applicant for JSAP Young Scientists Presentation Award

[9a-F211-9] Development of a Beam-Pattern Control System for Solid-State Laser Resonators
Using Deep Reinforcement Learning

〇Yuki Ikeya¹, Takashi Sekine¹, Endo Tsubasa², Toshio Otsu², Takaaki Morita¹, Yoshinori Tamaoki¹, Yoshinori Kato¹, Toshiyuki Kawashima¹, Yohei Kobayashi² (1.Hamamatsu Photonics K.K., 2.ISSP, Univ.Tokyo.)

Keywords:

liquid-crystal-on-silicon spatial light modulator,reinforcement learning,laser cavity

This study develops an AI-based system to automatically tune the beam pattern of a solid-state laser resonator. A camera module captures the beam image, and reinforcement learning updates the phase pattern on a programmable optical element (LCOS-SLM) based on the image features. By extending the control variables from 3 to 15 dimensions, we examine the effectiveness and challenges of autonomous control in a real system affected by measurement noise and limited trial speed.