| 학회 |
한국화학공학회 |
| 학술대회 |
2010년 봄 (04/22 ~ 04/23, 대구 EXCO) |
| 권호 |
16권 1호, p.173 |
| 발표분야 |
공정시스템 |
| 제목 |
Data-driven modeling of ammonia based CO2 capture processusing recursive kernel PLS |
| 초록 |
In recent years, aqueous ammonia has been received great attention as an effective absorbent in the post-combustion CO2 capture technology due to its high absorption capacity and low energy requirements. Despite these advantages, in the engineering point of view, the relatively complex process dynamics involved in the chemical reactions of H2O-CO2-NH3 mixture makes the assessment of current process condition very difficult during the operation and necessitates reliable analysis tool which can interpret the process behavior efficiently. Based on this consideration, a recursive kernel partial least squares (RKPLS) algorithm was newly proposed as an adaptive nonlinear multivariate statistical process control technique and applied to a pilot-scale plant for the monitoring and the prediction of overall process performance. The comparative studies were conducted by considering various kinds of statistical models and the results indicated that the RKPLS was superior to others in that it can efficiently take into account for the nonlinear and the time-varying characteristics of the given process. |
| 저자 |
장용수1, 이민우1, 이해우1, 안치규1, 김제영2, 한건우2, 박종문1
|
| 소속 |
1포항공과대, 2RIST |
| 키워드 |
CO2; absorption; ammonia; adaptive-KPLS
|
| E-Mail |
|
| 원문파일 |
초록 보기 |
