基于自适应算法WLSSVM的聚氯乙烯汽提工业模型  

Industrial Model of PVC Stripping Based on Adaptive Algorithm WLSSVM

作　　者：张斌[1] 李晓娜[1] 杨苗苗 ZHANG Bin;LI Xiao-na;YANG Miao-miao(Department of Basic Education,Henan Technical College of Construction,Zhengzhou 450064,China;General Education Center,Zhengzhou Business University,Zhengzhou 451200,China)

机构地区：[1]河南建筑职业技术学院基础教学部,河南郑州450064 [2]郑州商学院通识教育中心,河南郑州451200

出　　处：《塑料科技》2020年第8期84-87,共4页Plastics Science and Technology

摘　　要：聚氯乙烯(PVC)制备过程中,汽提塔温度控制精度决定了PVC产品的质量。利用加权最小二乘支持向量机(WLSSVM)建立了自适应WLSSVM的PVC汽提工业温度模型,对离群点分配较小的权值以提高模型精度。采用传统WLSSVM模型和自适应WLSSVM模型对汽提塔温度进行对比仿真实验。仿真实验结果表明,自适应WLSSVM模型相比于传统WLSSVM模型能够更好地预测PVC汽提工业中温度的变化。In the production process of polyvinyl chloride(PVC),the temperature control accuracy of the stripper determines the quality of the PVC product.A weighted least squares support vector machine(WLSSVM)was used to establish an adaptive WLSSVM PVC stripping industrial temperature model,and outliers were assigned smaller weights to improve the accuracy of the model.The traditional WLSSVM model and the adaptive WLSSVM model were used to compare and simulate the temperature of the stripper.The simulation experiment results showed that the adaptive WLSSVM model could better predict temperature changes in the PVC stripping industry than the traditional WLSSVM model.

关 键 词：汽提塔 WLSSVM 聚氯乙烯 温度控制 

分 类 号：TP273[自动化与计算机技术—检测技术与自动化装置]