基于支持向量机的高炉煤气利用率预测建模

煤气利用率是反映高炉能耗和平稳运行的重要指标。为了实现对高炉煤气利用率的准确预测,首先依据最大信息系数选择合适的输入参数,分别选取次于该状态参数时刻1 和2 h后的煤气利用率作为输出参数,并在建模之前对数据进行标准化处理。在此基础上建立基于支持向量回归(SVR)的高炉煤气利用率预测模型,并利用高炉的部分生产数据将该模型的预测结果与多层感知器(MLP)模型进行对比。最终预测结果表明,SVR模型在预测1和2 h后的煤气利用率时精确度更高,达到了更好的预测效果。

Forecast modeling of gas utilization rate of blast furnace based on support vector machine

Gas utilization rate (GUR) is an important indicator of blast furnace (BF) for reflecting the energy consumption and smooth operation of BF. In order to achieve accurate prediction of GUR, the appropriate input parameters are selected based on the maximum information coefficient. The gas utilization rates after one hour and two hours of the state parameter time are selected as output parameters, respectively. At the same time, it is essential for the model to standardize the data of BF. On this basis, a prediction model of GUR based on support vector regression (SVR) is established, and part of the production data of the BF is used to compare the prediction results of the model with those of the multi-layer perceptron (MLP) model. Final result demonstrates the SVR model is more accurate in predicting the GUR after one hour and two hours and achieves better prediction results.