论文部分内容阅读
稳定的炉温是保证高炉顺行、提高生铁质量和实现节能降耗的必要条件。针对高炉炉温与铁水硅质量分数及铁水温度之间存在的相关关系,利用某钢铁厂高炉的在线生产数据,构建一种双输出神经网络炉温预测模型,同时预测铁水硅质量分数和铁水温度,以便更加全面地预测炉温;同时,深入研究炉温时间序列对该模型性能的影响,确定较好的时间序列,建立基于时间序列的炉温预测模型。仿真结果表明,模型预测精度大大提高,可为高炉操作人员决策提供可靠依据。
Stable temperature is to ensure the smooth blast furnace to improve the quality of pig iron and achieve the necessary conditions for energy saving. According to the correlation between the blast furnace temperature and the mass fraction of molten silicon and the hot metal temperature, a dual output neural network temperature prediction model was constructed by using the online production data of a blast furnace in a steel plant. Simultaneously, the mass fraction of hot metal and the hot metal temperature , In order to predict the furnace temperature more comprehensively. At the same time, the influence of furnace temperature time series on the performance of the model was deeply studied, the better time series were determined, and the temperature prediction model based on time series was established. The simulation results show that the model prediction accuracy is greatly improved, which can provide a reliable basis for blast furnace operator decision-making.