CIESC Journal ›› 2014, Vol. 65 ›› Issue (12): 4883-4889.DOI: 10.3969/j.issn.0438-1157.2014.12.032

Previous Articles     Next Articles

Soft-sensor modeling for ethylene distillation product quality based on vector projection metabolism support vector machine

ZHENG Boyuan1, SU Chengli1, LI Ping1, SU Shengjiao2   

  1. 1. School of Information and Control Engineering, Liaoning Shihua University, Fushun 113001, Liaoning, China;
    2. Fushun Petrochemical Branch Company, PetroChina, Fushun 113004, Liaoning, China
  • Received:2014-05-20 Revised:2014-08-13 Online:2014-12-05 Published:2014-12-05
  • Supported by:

    supported by the National Natural Science Foundation of China (61203021) and the Scientific and Technological Project of Liaoning Province (2011216011).

基于向量投影的代谢支持向量机乙烯精馏产品质量软测量建模

郑博元1, 苏成利1, 李平1, 苏胜蛟2   

  1. 1. 辽宁石油化工大学信息与控制工程学院, 辽宁 抚顺 113001;
    2. 中国石油天然气股份有限公司抚顺石化分公司, 辽宁 抚顺 113004
  • 通讯作者: 苏成利
  • 基金资助:

    国家自然科学基金项目(61203021);辽宁省科技攻关项目(2011216011).

Abstract: A metabolism support vector machine (SVM) based modeling method was proposed to solve the problems that slowed down computing speed and caused poor stability during SVM incremental learning. Firstly, vector projection algorithm was used to pre-extract training samples, in order to reduce the number of samples and improve SVM modeling speed. Secondly, a new sample "metabolism"principle was pulled into the SVM incremental learning process, for addressing "explosion" of training samples number which was caused by adding new samples continuously. Finally, the vector projection metabolism SVM was utilized in ethylene distillation product quality soft-sensor modeling. The experiment results showed that vector projection metabolism SVM had better prediction result than SVM and LSSVM.

Key words: dynamic modeling, prediction, model, ethylene distillation, metabolism support vector machine, vector projection, soft-sensor

摘要: 针对支持向量机(SVM)增量学习过程中易出现计算速度慢、稳定性差的缺陷,提出了一种基于向量投影的代谢支持向量机建模方法.该方法首先运用向量投影算法对训练样本进行预选取来减少样本数量,提高SVM建模速度.然后将新增样本"代谢"原则引入SVM增量学习过程中,以解决因新增样本不断加入而导致训练样本数量"爆炸"的问题.最后将该方法用于乙烯精馏产品质量软测量建模,实验结果表明,与传统SVM和最小二乘支持向量机(LSSVM)相比,向量投影的代谢SVM具有更好的预测结果.

关键词: 动态建模, 预测, 模型, 乙烯精馏, 代谢支持向量机, 向量投影, 软测量

CLC Number: