基于灰箱模型的LF精炼终点温度预测

以国内某钢厂一30 t LF精炼炉为研究对象，通过建立由传热机理模型和黑箱模型相结合的灰箱模型对LF精炼终点温度进行了预测。首先根据能量守恒定律建立了传热机理模型。针对包衬耐材的蓄热以及合金的热效应难以精确计算的问题，采用偏最小二乘黑箱模型对这一部分温度进行了处理，最后将两种模型相结合综合预测了LF钢包精炼的终点温度。结果表明，偏最小二乘法在预测包衬的耐材蓄热和合金热方面的温度误差在±5℃以内的命中率达到97%以上，总的灰箱模型预测LF精炼终点温度误差在±5、±8、±10℃以内的命中率分别达到88%、96%和99%，模型具有较高的预测精度。研究可为该钢厂的LF精炼工艺提供指导。

End-point temperature prediction of molten steel in LF based on gray-box model

Taking a 30 ton of LF as an investigated object, a gray-box model to predict the end-point temperature of molten steel in LF based on the mechanism model and the black-box model was established. The heat transfer mechanism model was firstly estab-lished according to the energy conservation law. In addition, the black-box model based on partial least squares was adopted to solve the problem that the regeneration of the ladle lining and thermal effects of the alloys are difficult to accurately calculate. Finally, the two models were combined to predict the end-point temperature of molten steel in LF. Simulation results show that the hit rate within the error range of ±5℃ by partial least squares for the regeneration of ladle lining refractory and thermal effects of the alloys reaches more than 97%, and the total hit rate within the error ranges of ±5, ±8 and ±10℃ by gray-box model reaches 88%, 96% and 99%, respectively. This means the grey-box model can accurately predict the end-point temperature of LF, which provides the guid-ance for the LF refining process in this factory.