유동해석을 통한 스마트 제어밸브 학습 및 성능평가용 프로파일 모델링 = Flow Analysis and Profile Modeling for Machine Learning and Performance Evaluation of Control Valve

다국어 초록 (Multilingual Abstract)

To evaluate performance of DB based control valve algorithm, CFD analysis was performed to generate any imaginary sensing data. Pressure loss coefficient(K) and flow-rate coefficient(Cv) of globe valve were acquired using CFD analysis results. Feasible 3-variable fitting(Dose-Response-2D relationship) was peformed to generate relationship with valve travel, pressure and flow-rate. Arbitrary valve travel profile could be acquired using relationship generated. The profile for learning and evaluation of AI valve algorithm can be plotted also through fitting method above.

참고문헌 (Reference)

1 Stephen Marsland, "Machine Learning : An Algorithmic Perspective" CRC 225-230, 2016

2 Jo, S.J, "Industrial Application of Machine Learning(A.I.)" 23 (23): 34-38, 2016

3 Etsuji Nakai, "IT Engineer No Tame No Kikai Gakushu Riron Nyumon" Gijyutsu-Hyoron 44-73, 2016

4 Crump, K.S, "Fundamental carcinogenic processes and their implications for low dose risk assessment" 36 (36): 2973-2979, 1976

5 "Dario Cambié"

6 Lee, Y.H, "CFD Study of 3-way Temperature Control Valve of Cooling Line of Emergency Generator for Nuclear Power Plant" 229-230, 2013

7 Ko, J.W, "A Study of the valve flow analysis for deriving parallel valve driving method" 225-226, 2013