DETECTION OF MALFUNCTIONS AND ABNORMAL OPERATING CONDITIONS OF A COAL MILL

摘要: Coal mill malfunctions are some of the most common causes failing to keep power plant crucial operating parameters or even unplanned shutdowns. Therefore, an algorithm has been developed that enable on-line detection abnormal conditions and mill. Based on calculated diagnostic signals defined thresholds, this informs about conditions. Diagnostic represent difference between measured modeled value of two selected parameters. Models motor current outlet temperature pulverized fuel were based linear regression theory. Various data analyses feature selection procedures have performed obtain best possible model. The model compared with two alternative models. first was artificial neural network second physical equations use a genetic determine unknown coefficients. validation carried out historical containing values from 10 months operation. Historical downloaded DCS system 200 MW coal-fired plant. In set failures occurred six times causing Tests show can be successfully used detect certain mill, such as: feeder blockage, lack coal, overload.