基于GA-SVM的烟气含氧量软测量  

Soft Sensing of Oxygen Content in Flue Gas Based on GA-SVM

作　　者：王永智 付忠广[1] 刘炳含 王鹏凯 

机构地区：[1]华北电力大学

出　　处：《化工自动化及仪表》2017年第10期937-939,共3页Control and Instruments in Chemical Industry

摘　　要：针对电站锅炉烟气含氧量传统硬件测量方法成本昂贵、使用寿命短等缺陷,提出一种基于支持向量机的软测量方法。首先结合机理分析和数据相关性分析选取相关过程参数作为模型输入参数,使用遗传算法对支持向量机进行参数寻优,构建基于遗传算法参数优化的支持向量机(GA-SVM)软测量模型。实验结果表明:该模型能较好地反映烟气含氧量的变化趋势。Considering higher cost and shorter service life of traditional measuring methods for the oxygen con- tent in boiler flue gas of the power plant, a soft sensing method based on support vector machine（SVM） was proposed. It has mechanism analysis and data correlation analysis combined to choose relevant process parame- ters as the model input parameters and has the genetic algorithm（GA） employed to optimize the parameters of support vector machine and to establish the soft sensing model based on the support vector machine that param- eters are optimized by genetic algorithm（GA-SVM）. The experimental result shows that, this model can well reflect the change trend of oxygen content in flue gas.

关 键 词：软测量 电站锅炉 烟气含氧量 GA-SVM软测量模型 

分 类 号：TH83[机械工程—仪器科学与技术]