基于支持向量机插值的铜电解液成分检测-JournalofNortheastern.PDF

第29 卷第2期 东 北 大 学 学 报 ( 自 然 科 学 版 ) Vol29, No. 2 2 00 8 年 2 月 Journal of Northeastern U niversit ( Natural Science) Feb. 2 0 0 8 1, 2 1 1, 2 袁 平 , 李 妍 , 毛志忠 ( 1. , 110004; 2. , 110004) : , , , , , , , , , : ; ; ; ; : TP 273. 1 : A : 1005-3026( 2008) 02-0 174-04 Detection Based on Support Vector Machine Interpolation for Copper Electrolytic Composition 1, 2 1 1, 2 YUA N Ping , LI Yan , MA O Zhi-z hong ( 1. School of Information Science Engineering , Northeastern Universit , Shen ang 110004, China; 2. Ke Laborator of Integrated Automation of Process Industr , Ministr of Education, Northeastern Universit , Shen ang 110004, China. Correspondent: LI Y an, E-mail: an0317 @ sina. com) Abstract : In the process of electrol tic copper production the contents of copper ions and acid ions can onl be detected once each da through tests, and the are inadequate to show ing the regularit of variation of copper electrol te composition. It w as found that the values for conventional interpolation often differ greatl from the values obtained actuall in site. To solve the problem, a soft measurement model is advanced considering the influencing factors on measurement with relevant data forecasted efficientl . The support vector machine is introduced to improve linear interpolation because of its advantages in developing the soft measurement model. And a method is given to determine the weights for the values estimated b the soft me