PURPOSE:To be able to identify burning condition exactly and reduce the generated quantity of exhaust gas with the adjustment of burner operation by storing previously relations between fire flame images and extracted features quantity, and burning condition, the generated quantity of exhaust gas, and control amount. CONSTITUTION:Image input 1 ensures presence safety of flame, check and uniforming of safety by judgement on presence of flame, average processing of images etc. Image processor 2 compensates for the rotation of flame images due to torsion of image fiber etc. and the parallel movement of coordinates due to the cut-out of images and eliminates noise components included in flame images. Feature quantity extraction processor 3 extracts feature quantity by using area calculation, calculation of gravity center, calculation of circumference, distance calculation etc. Knowledge processor 4 reasons burning condition by using relations among a flame and its feature quantity and the quantity of exhaust gas components, fuel quantity/air quantity, secondary air quantity/third air quantity etc. previously stored as information. Conclusion indicator 5 outputs each processed result on a display unit when specified as the need arises. By said processing, burning condition during the operation time of a boiler is identified quantitatively and also qualitatively by reasoning using knowledge processing.
展开▼
