Abstract:Under the influence of dynamic factors such as belt input raw materials and coke falling off, the melting pool temperature, matte grade and slag type in the furnace are prone to change rapidly. If the furnace condition is not adjusted in time, it is easy to cause crust in the furnace, and even dead furnace may occur in serious cases. In view of this, the feasibility of rapid on-line composition detection of matte and slag by laser induced breakdown spectroscopy (LIBS) is explored, and then the detection data is manually transmitted to Metcal metallurgical digital simulation software for feedback and guidance of production. Through the long-term operation practice of the LIBS melt online detection equipment, the reliability and accuracy of its detection are verified. Based on the X-ray fluorescence spectrometer(XRF) detection results, the detection deviation of LIBS melt is ±1.0% to 96.2%, and the solid detection deviation is ±1.0% to 81.7%. Basically, it can be realized that LIBS can replace XRF detection. Based on the high real-time online testing results, the deviation between the average actual oxygen consumption of 15 furnaces and the LIBS guidance digital oxygen consumption was reduced from 7.65% to 2.77%, so as to accurately regulate the furnace condition control parameters, providing an important guarantee for the stable production and energy saving of furnaces.