KPCA和LSSVM在SCR脱硝反应器入口NO_x含量软测量中的应用

SCR脱硝反应器入口NO_x含量及时、准确地测量,对于精确调节喷氨量,控制NO_x排放至关重要。针对NO_x气体分析仪测量滞后的问题,提出了基于KPCA和LSSVM的软测量模型。根据某电厂采集的数据样本规模较大的情况,为了提高NO_x软测量模型的精度,该文首先进行基于相似度函数的样本优选,减少样本冗余信息,简化样本模型。然后对于选取的与反应器入口NO_x有关的18个辅助变量进行核主元分析(KPCA),对样本进行特征提取,降低样本维数,以此作为最小二乘支持向量机(LSSVM)软测量模型的输入,提高软测量模型的精度,对反应器入口NO_x含量的实时、准确测量提供一定的理论依据,为提高反应器脱硝效率打下良好基础。

The application of KPCA and LSSVM in the soft-sensing of SCR denitration reactor inlet NOx content

The timely, accurate measurement of SCR denitration reactor inlet NOx content is very important to accurately adjust the amount of ammonia spray and the NOx emission control. According to the serious delay of the NOx emission analyzer which used in the power plant, a soft-sensing based on Kernel Principal Component Analysis( KPCA) and Least Support Vector Machine( LSSVM) is proposed. The original data is acquired from a certain power plant, and the sample size is very large. The KPCA method is not suitable for the analysis of a large number of samples. So firstly use the method of similarity function optimization to simplify the input samples and reduce the redundant information. In order to improve the timely, accuracy of NOx -content, 18 auxiliary parameters that connected to NOx were analyzed. The ( LSSVM) with the KPCA was used to analyze the related parameters and established the soft measurement model.