Presentation Information
[17p-PA2-3]Laser beam stabilization using data-driven feedforward control based on deep learning
〇Shogo Takahashi1, Yoshio Hayasaki1, Satoshi Hasegawa1 (1.Utsunomiya Univ.)
Keywords:
deep learning
Laser technology is widely used in various fields such as materials processing and precision measurement, where beam stability is a critical factor determining processing quality and measurement accuracy. However, external disturbances such as mechanical vibrations and temperature fluctuations cause intensity variations and beam pointing instability. In this study, we propose a real-time laser beam stabilization method that predicts future beam fluctuations using a recurrent neural network (RNN) and performs feedforward compensation with a spatial light modulator(SLM), enabling effective suppression even for high-speed fluctuations.