| Autor: |
Virgiawan, Akbar Rachmad, Apriyanto, Adhi Eko, Pariaman, Henry |
| Předmět: |
|
| Zdroj: |
AIP Conference Proceedings; 2020, Vol. 2217 Issue 1, p030139-1-030139-9, 9p, 3 Diagrams, 9 Charts, 2 Graphs |
| Abstrakt: |
This paper proposes using Big Data and Similarity Based-Modelling (SBM) to identify early alert in online monitoring of Paiton Coal-Fired Power Plant operating system. Similarity-based modeling is a nonparametric modeling technique that uses the similarity of a query vector with exemplar vectors to infer the model’s response. Big Data and SBM has been successfully used for anomaly detection on industrial applications. This method use historical operational data that has been filtered to eliminate outlier data/bad data. That filtered data become predicted data. After that, come actual data from Site and it will be compare with predicted data using SBM Method to search the condition of predicted data that almost the same with the actual data. If there are difference and at a certain time range, it will show the Early Alert. Early Alert is the indication of disturbance and has five (5) priority. The higher indication of disturbance, the priority also higher. With this method, Forced Outage can be reduced so that Power Plant’s Reliability can be good. [ABSTRACT FROM AUTHOR] |
| Databáze: |
Complementary Index |
| Externí odkaz: |
|
