基于KPCA的SBR过程监视  

Monitoring of SBR Process Using Kernel Principal Component Analysis

作　　者：樊立萍[1] 于海斌[2] 袁德成[1] 徐阳[1] 

机构地区：[1]沈阳化工学院自动化系,沈阳110142 [2]中国科学院沈阳自动化研究所,沈阳110016

出　　处：《仪器仪表学报》2006年第3期249-253,共5页Chinese Journal of Scientific Instrument

摘　　要：序批式反应器生化污水处理系统(SBR)具有复杂的生化反应机理,其固有的严重非线性、持续时间有限、非稳态运行等给其过程监视带来特殊困难。核主元分析(KPCA)方法通过集成算子与非线性核函数计算高维特性空间的主元成分,有效捕捉过程变量中的非线性关系。将KPCA技巧应用到序批式反应器生化污水处理系统,建立了基于KPCA的SBR污水处理过程在线监视策略。在监视暴风雨事件等典型的SBR过程异常状态时,统计指标变化灵敏,诊断及时。与线性PCA相比,显示出更高的过程监视性能。Sequencing batch reactor （SBR） for biological wastewater treatment has complex biochemical mechanism. Its inherent high nonlinearity, limited duration and astable running bring about exceptional difficulties in process monitoring. Kernel principal component analysis （KPCA） can compute principal components in high-dimensional feature space by means of integral operators and nonlinear kernel functions. It can effectively capture the nonlinear relationship in the process variables. KPCA trick was applied to the SBR biological wastewater treatment processes and an on-line process monitoring strategy for SBR was proposed. In monitoring the abnormal conditions of SBR process, such as rainstorm events, the statistical indexes vary sensitively and give timely diagnose. In comparison to PCA, KPCA showed superior process monitoring performance.

关 键 词：核主元分析(KPCA) 序批式反应器(SBR) 生化污水处理 监视 

分 类 号：TP206.3[自动化与计算机技术—检测技术与自动化装置]