基于数据挖掘技术的SCR脱硝反应器入口NO_(x)软测量方法

Soft-Sensing Method for Inlet NO_(x) of SCR Denitration Reactor Based on Data Mining Technology

针对选择性催化还原系统(SCR)严重的滞后性特点,以电厂实际运行数据为基础数据,分别采用灰色系统GM(1,N)模型和BP神经网络进行试验数据的挖掘,寻找试验过程的重要控制参数,并对其进行入口氮氧化物(NO_(x))预测分析。通过比较两个模型的预测结果可知,BP神经网络模型比灰色系统GM(1,N)模型具有更高的预测精度,更适合于SCR入口NO_(x)的预测。

In view of the serious nonlinearity,hysteresis and time variability of the selective catalytic reduction system(SCR),the gray system GM(1,N)model and the BP neural network are used to test the actual operating data of the power plant,looking for important control parameters of the test process and conducting the inlet nitrogen oxide(NO_(x))prediction analysis.By comparing the prediction results of the two models,it can be seen that the BP neural network model has higher prediction accuracy than the gray system GM(1,N)model and is more suitable for prediction of SCR inlet NO_(x).