考察了影响LF炉钢水温度的因素。从能量平衡的角度出发,将整个钢包体系作为1个系统,确定加热功率、钢水质量、钢包温度、包龄、渣厚、氩气吹入量、时段7个主要因素作为网络的输入量,应用BP神经元网络进行初步预报,再根据专家工艺知识对一些特殊情况进行修正。使用本方法可减少点测次数,获得连续的钢水温度信息,降低炼钢成本,提高质量。预报误差在±10℃以内。 The factors affecting the temperature of LF molten steel were investigated. From the energy balance point of view, the entire ladle system as a system to determine the heating power, molten steel quality, ladle temperature, age, slag thickness, argon inflow, time 7 main factors as the network input, application BP neural network for the initial forecast, and then based on expert knowledge of some special cases to be amended. Using this method can reduce the number of spot measurement, obtain continuous molten steel temperature information, reduce the cost of steelmaking and improve the quality. Forecast error within ± 10 ℃.