Judging Furnace Outfire Based on Furnace Operation Conditions Support Vector Machine

摘要

A method of judging furnace burning outfire based on flame image information was studied. Three feature vectors were extracted from furnace flame images. A method of using Support Vector Machine combined with furnace operation conditions to judge furnace outfire was designed. According to signals of coal mill outage and coal valve switch, burner outfire was judged when powder burner off. And combustion states were classified by Support Vector Machine when powder is being sprayed. The method is simple, convenient and practical and has high correct classification rate.