| 학회 |
한국공업화학회 |
| 학술대회 |
2022년 봄 (05/11 ~ 05/13, 제주국제컨벤션센터(ICC JEJU)) |
| 권호 |
26권 1호 |
| 발표분야 |
포스터-화학공정 |
| 제목 |
Multi-objective optimization of CO2 emission and thermal efficiency for steam reforming process using deep neural network |
| 초록 |
Hydrogen is mainly produced through steam reforming that produces CO₂ as a by-product. Therefore, it is important to optimize operating conditions for maximizing thermal efficiency and minimizing CO₂ emission simultaneously. This study focuses on multi-objective optimization considering the two objectives mentioned above. A hybrid deep neural network (DNN)-based model is developed to increase the robustness of multi-objective optimization. The developed hybrid DNN model is integrated with the multi-objective particle swarm optimization algorithm that performs Pareto dominance-based multi-objective optimization. As a result, Pareto optimal solutions are obtained with a thermal efficiency distribution of 77.5 to 87.0% and CO₂ emission of 577.9 to 597.6 t/y. Moreover, the Pareto-optimal front was evaluated to provide various representative solutions to decision-makers. An implication of this study is enabling efficient and flexible process operation according to different requirements. |
| 저자 |
이재원1, 홍석영2, 조형태1, 박진우3, 이인규4, 김정환1
|
| 소속 |
1한국생산기술(연), 2연세대, 3공주대, 4부산대 |
| 키워드 |
steam reforming; hydrogen; CO2 emission; deep neural network; multi-objective optimization
|
| E-Mail |
|
