Intelligent methods for process control and diagnostics of a mill fan system

Abstract

Abstract The intelligent methods for process control and diagnostics of the mill fan system is an established field of scientific and applied investigations. In the present paper several types of process control approaches with different structures are considered. In order to choose the most efficient one, comparative analysis is carried out. The mill fans are a basic element of the dust-preparing systems of steam generators with direct breathing of the coal dust in the furnace chamber. Such generators in Bulgaria are the ones in Maritsa East 2 Thermal Power Plant, in Maritsa East 3 Thermal Power Plant and also in Bobov Dol Thermal Power Plant. The subject of this research is a device from Maritsa East 2 Thermal Power Plant. This is the largest thermal power plant on the Balkan Peninsula. Standard statistical and probabilistic (Bayesian) approaches for diagnostics are inapplicable to estimate the mill fan technical state due to non-stationarity, non-ergodicity and the significant noise level. The possibility to predict eventual damages or wearing out without switching off the device is significant for providing faultless and reliable work, avoiding the losses caused by planned maintenance.