基于WinCC的SO_2转化率在线软预测及校正研究

Research on Online MLR Soft-forecast and Adjustment of SO_2 Conversion Ratio Based on WinCC

针对冶炼烟气制酸中SO2转化率很难直接在线测量,采用多元线性回归(MLR)方法建立SO2转化率的软测量模型,并对MLR模型进行了校正研究,以提高SO2转化率的预测精度。为了便于工程应用,以目前国内外主流的西门子Win CC(Windows Control Center)为监控平台,采用VBS(Microsoft Visual Basic Script Editon)脚本将SO2转化率MLR软测量模型及其校正算法嵌入Win CC中,实现了及时在线预测SO2转化率的目的,为指导生产提供了依据。

To solve the problem that the SO2 conversion ratio of smelting off-gas acid is difficult to be directly measured on-line, this paper establishes a soft measurement model of the SO2 conversion ratio by adopting the Multiple Linear Regression （MLR） method. In addition, to improve the prediction accuracy of the ratio, researches have been done to adjust the MLR model. Furthermore, in order to facilitate engineering applications, the mainstream Siemens WinCC is used as a monitoring platform, the VBS script is applied to insert the MLR soft measurement model of the SO2 conversion ratio and its correction algorithm into WinCC to realize timely online prediction of SO2 conversion ratio, which provides guidance for the real-time production.