基于BP神经网络的LF精炼终点温度预测模型

Prediction Model of LF Refining Endpoint Temperature Based on BP Neural Network

通过对LF精炼过程的研究找出了对LF精炼终点温度起主要作用的工艺参数,应用BP神经网络建立了终点拟合仿真函数模型。应用现场实际测量结果与模型预测结果进行对比,温度误差小于5℃的炉次为总炉次的96.5%,温度误差最大值为8℃。

Through the study of LF refining process,this paper finds out the process parameters that play a major role in the end point temperature of LF refining,and establishes a simulation function model of end point fitting by using BP neural network.By comparing the measured results with the predicted results of the model,96.5%of the total number of stoves with a temperature error less than 5℃and the maximum temperature error of 8℃were obtained.