基于SVM算法的SCR催化剂使用寿命多层预测模型

Multi layer prediction model of SCR catalyst service life based on SVM algorithm

文章使用多层预测模型结合SVM算法实现对于SCR催化剂的寿命预测,以某厂SCR装置的运行数据为依据,研究对象为不同操作条件下的钒基SCR催化剂。利用灰色关联度分析确定变量,建立关于催化剂相对活性的预测模型,通过催化剂相对活性的变化来预测催化剂的寿命。将运行数据与模型预测模拟数据进行对比,证明模型拟合程度较好。

In the paper,using multi-layer prediction model combined with SVM algorithm to realize the life prediction of SCR catalyst,according to the operation data of a power plant,the research object is the niobium-based SCR catalyst under different operating conditions.The variables are determined by gray correlation analysis,and the prediction model of the relative activity of the catalyst is established.The life of the catalyst is predicted by the change of the relative activity of the catalyst.Compares the running data with the model prediction simulation data,and the model fits better.