| 학회 |
한국화학공학회 |
| 학술대회 |
2020년 가을 (10/14 ~ 10/16, e-컨퍼런스) |
| 권호 |
26권 1호, p.880 |
| 발표분야 |
화학공정안전 |
| 제목 |
비용과 안전을 고려한 보일러 공정의 최적화 운전 조건 탐색 |
| 초록 |
In case of a boiler process of a pulp mill, it is impossible to use conventional modeling software. To perform a modeling for a boiler process in this study, support vector regression is introduced to mimic a dynamic model of a boiler. To find the optimal operational condition under various uncertainties, Monte-Carlo based sample average approximation is introduced to generate a stochastic optimization problem. In addition, a particle swarm optimization technique, which is one of sampling approaches and does not need the derivatives of equations, is used to find a stochastic optimal solution, since our models based on a support vector regression is not able to provide their differential equations. To verify the performance of a stochastic optimal solution, the value of the stochastic solution is calculated and the result shows that the performance of a stochastic solution provides better performance than that of a deterministic solution. The proposed study can be easily to black box models. |
| 저자 |
이창준1, Gustavo Matheus de Almeida2, Song Won Park3
|
| 소속 |
1부경대, 2Federal Univ. of Minas Gerais, 3Univ. of São Paulo |
| 키워드 |
|
| E-Mail |
|
| 원문파일 |
초록 보기 |
