OMFP: An Approach for Online Mass Flow Prediction in CFB Boilers

Authors:
Indrė Žliobaitė;Jorn Bakker;Mykola Pechenizkiy
Affiliations:
Department of Computer Science, TU Eindhoven, Eindhoven, The Netherlands NL-5600;Department of Computer Science, TU Eindhoven, Eindhoven, The Netherlands NL-5600;Department of Computer Science, TU Eindhoven, Eindhoven, The Netherlands NL-5600 and Dept. of MIT, U. Jyväskylä, Finland FIN-40014
Venue:
DS '09 Proceedings of the 12th International Conference on Discovery Science
Year:
2009

Citing 6
Cited 1

Topics in matrix analysis

Topics in matrix analysis
Adapting to Drift in Continuous Domains (Extended Abstract)

ECML '95 Proceedings of the 8th European Conference on Machine Learning
Anomaly detection: A survey

ACM Computing Surveys (CSUR)
Handling outliers and concept drift in online mass flow prediction in CFB boilers

Proceedings of the Third International Workshop on Knowledge Discovery from Sensor Data
Learning with local drift detection

ADMA'06 Proceedings of the Second international conference on Advanced Data Mining and Applications
Mining temporal data: a coal-fired boiler case study

KES'05 Proceedings of the 9th international conference on Knowledge-Based Intelligent Information and Engineering Systems - Volume Part III

Quantile index for gradual and abrupt change detection from CFB boiler sensor data in online settings

Proceedings of the Sixth International Workshop on Knowledge Discovery from Sensor Data

Quantified Score

Hi-index	0.00

Visualization

Abstract

Fuel feeding and inhomogeneity of fuel typically cause process fluctuations in the circulating fluidized bed (CFB) boilers. If control systems fail to compensate the fluctuations, the whole plant will suffer from fluctuations that are reinforced by the closed-loop controls. Accurate estimates of fuel consumption among other factors are needed for control systems operation. In this paper we address a problem of online mass flow prediction. Particularly, we consider the problems of (1) constructing the ground truth , (2) handling noise and abrupt concept drift, and (3) learning an accurate predictor. Last but not least we emphasize the importance of having the domain knowledge concerning the considered case. We demonstrate the performance of OMPF using real data sets collected from the experimental CFB boiler.