基于最小二乘支持向量机的软测量方法在精对苯二甲酸生产中的应用  

Soft Sensor Modeling Based on Least Square Support Vector Machine and Its Applications to Control Pure Terephthalic Acid

作　　者：王清培[1] 刘瑞兰[2] 

机构地区：[1]北京大学遥感与地理信息系统研究所,北京100871 [2]南京邮电大学自动化学院,南京210003

出　　处：《化工自动化及仪表》2010年第7期32-34,39,共4页Control and Instruments in Chemical Industry

摘　　要：研究了一种基于改进型的最小二乘支持向量机的软测量方法在精对苯二甲酸(PTA)生产中晶体粒径大小估计中的应用。对最小二乘支持向量机采取了稀疏化与实时滚动改进,改进后的算法样本数量大大降低,同时保持了原有水平的预测能力。结果表明该方法对精对苯二甲酸的颗粒具有较高的预测能力。A soft sensor method which based on least square support vector machine（LS-SVM） was introduced to the pure terephthalic acid（PTA） estimation.The sparseness of LS-SVM was also be added in order to improve the algorithm efficiency by removing the vector which had small impact to the result.The real time pattern with small samples was introduced to make the algorithm keep both high estimate efficiency and execution efficiency.The result shows that this algorithm can estimate the PTA at a high quality.

关 键 词：支持向量机 软测量 精对苯二甲酸 实时估计 

分 类 号：TP181[自动化与计算机技术—控制理论与控制工程] TP273[自动化与计算机技术—控制科学与工程]