과제정보
This work was supported by a Grant (07UrbanRenessanceB03) from High-Tech Urban Development Program funded by the Korean Ministry of Land, Transport, and Maritime Affairs.
국제학술발표논문집 (International conference on construction engineering and project management)
- The 3th International Conference on Construction Engineering and Project Management
- /
- Pages.388-392
- /
- 2009
- /
- 2508-9048(eISSN)
PREDICTION OF FAULT TREND IN A LNG PLANT USING WAVELET TRANSFORM AND ARIMA MODEL
- Yeonjong Ju (School of Civil and Environmental Engineering, College of Engineering, Yonsei University ) ;
- Changyoon Kim (School of Civil and Environmental Engineering, College of Engineering, Yonsei University) ;
- Hyoungkwan Kim (School of Civil and Environmental Engineering, College of Engineering, Yonsei University)
- 발행 : 2009.05.27
초록
Operation of LNG (Liquefied Natural Gas) plants requires an effective maintenance strategy. To this end, the long-term and short-term trend of faults, such as mechanical and electrical troubles, should be identified so as to take proactive approach for ensuring the smooth and productive operation. However, it is not an easy task to predict the fault trend in LNG plants. Many variables and unexpected conditions make it quite difficult for the facility manager to be well prepared for future faulty conditions. This paper presents a model to predict the fault trend in a LNG plant. ARIMA (Auto-Regressive Integrated Moving Average) model is combined with Wavelet Transform to enhance the prediction capability of the proposed model. Test results show the potential of the proposed model for the preventive maintenance strategy.