针对采用吸收法捕集CO2的过程中存在着成分难以准确在线测量的问题,设计了工业化应用最广泛的乙醇胺(MEA)化学吸收过程的软测量系统。系统使用红外传感器,通过基于改进Powell鲁棒支持向量机(IP-RSVM)的方法对吸收溶液的离子组分分布进行了软测量,其结果与核磁共振(NMR)测量所得的数据符合较好,均方误差为3.1692×10-7,最大绝对误差绝对值为0.0010,最大相对误差为3.54%。研究结果表明,在CO回收的软测量系统中,使用基于IP-RSVM的红外软测量具有良好应用前景。
In order to overcome the difficulty of online measuring the main components in the process of capturing CO2 from the flue gas of a power plant based on absorption method,a soft-sensing system was designed for the process of recovering CO2 by MEA chemical absorption.Based on the principle of improved Powell robust support vector machine (IP-RSVM),the soft measurement of ionic species distribution in the absorption solution was carried out by using infrared sensors,and the results agreed well with the data obtained by nuclear magnetic resonance (NMR) measurement.The mean square error was 3.1692×10-7,the maximum value of absolute error was 0.0010,and the maximum value of relative error was 3.54%.It showed that the infrared soft measurement based on IP-RSVM could be used in the soft-sensing system of capturing CO2 effectively.