Endpoint prediction of BOF by flame spectrum and furnace mouth image based on fuzzy support vector machine

摘要

Aiming at the dynamic endpoint prediction of the basic oxygen furnace (BOF), a new method based on features of the flame spectrum and furnace mouth with fuzzy support vector machine (FSVM) is proposed for converters whose vessel mouths are not stationary. A non-contact system is designed for light radiation spectrum and image acquisition of furnace mouth. Parameters characterizing the overall spectrum fitted by Gauss function and emission peaks are extracted from spectrum and state parameters of furnace mouth are extracted from image respectively. The extracted parameters are used as inputs of the FSVM for modeling. The experimental results show that the proposed method has better recognition accuracy than the method of SVM with only spectrum even the furnace mouth changes dynamically.