Development of estimation measure for BWR LOCA break condition using machine learning technology

Daisuke Fujiwara

Presentation information

Oral presentation

IV. Nuclear Plant Technologies » 402-1 Nuclear Safety Engineering (Safety Design, Safety Assessment/Analysis, Management)

[3L10-14] Safety Analysis

Fri. Mar 22, 2019 2:45 PM - 4:10 PM Room L (Common Education Bildg. 2 3F No.36)

Chair:Akifumi Yamaji(Waseda Univ.)

3:15 PM - 3:30 PM

[3L12] Development of estimation measure for BWR LOCA break condition using machine learning technology

*Daisuke Fujiwara¹, Hiroshi Shirai¹, Ryunosuke Yamaguchi¹ (1. TEPSYS)

Keywords:LOCA break condition, Emergency responce system, Machine learning

Estimation measure for BWR LOCA condition has been developed using machine learning technology. This measure utilizes observable parameters in BWR as learning data and estimate LOCA break area and elevation. Prediction accuracy and future issues will be reported and prospection of incorporation into emergency responce system will be introduced.