在线监测燃煤锅炉NOx排放的自适应支持向量机模型

提出了一种基于改进的在线支持向量机自适应建模方法,并应用于电站锅炉NOx排放连续监测和特性分析.对常规在线支持向量机方法进行了改进,提出了新的样本剔除规则,保证了训练集内样本分布的均匀性.通过该改进方法对基于试验数据的常规向量机模型预测余差进行了连续估计,并预估煤质等因素引起的NOx排放特性的变化,从而补偿了实际工况与试验工况的差别,以便正确给出锅炉NOx排放特性.

An Adaptive Support Vector Machine Model for the On-line Monitoring of Boiler NOx Emissions in Coal-fired Boiler

An adaptive modeling method,based on a modified accurate on-line support vector regression(AOSVR),is proposed for continuous monitoring of boiler NO_x emissions in coal-fired boiler.A modified criterion for selection of the unwanted trained sample was introduced to maintain the uniformly distribution of samples in training set.The modified AOSVR was used to adaptively predict residuals of the conventional SVR model and NO_x emissions change caused by the variation of coal quality etc.It can compensate the deviation between test condition and actual condition,and give the valid prediction of NO_x emissions.