污水处理过程生化需氧量智能集成软测量模型  

Biochemical Oxygen Demand Intelligent Integrated Soft-sensing Model for Sewage Treatment Process

作　　者：宋剑杰[1] 

机构地区：[1]湖南科技职业学院电子信息系,湖南长沙410004

出　　处：《自动化仪表》2011年第10期42-45,共4页Process Automation Instrumentation

摘　　要：针对污水处理过程中生化需氧量(BOD)难以在线检测的问题,提出一种基于满意聚类和改进SVM的BOD智能集成软测量建模方法。采用满意聚类算法对预处理后的数据样本进行聚类分析,利用改进SVM为每个聚类样本建立BOD子预测模型,并采用模糊组合方法获得最终的BOD预测值。仿真试验表明,该方法预测精度高,能够满足污水处理过程对BOD检测的实际要求。Aiming at the difficulty of online detection for biochemical oxygen demand（BOD） parameter in sewage treatment process,the BOD intelligent integrated soft-sensing modeling method based on satisfactory clustering and improved SVM is proposed.First,the satisfactory clustering is used to do the clustering analysis on the pretreatment sample data;then the improved SVM is utilized to establish the BOD sub-prediction model for each cluster;finally,the predictive value of BOD is obtained by adopting fuzzy combination method.The method features higher predictive accuracy and meets the practical requirement for BOD detection in sewage treatment process.

关 键 词：生化需氧量 改进SVM 污水处理 软测量 建模 

分 类 号：TP29[自动化与计算机技术—检测技术与自动化装置]