工业共沸精馏塔软测量建模方法的研究与应用

针对某一工业共沸精馏塔成分估计问题,利用基于支持向量机技术的软测量建模方法,建立了恰当的工业软测量模型。利用滑动时间窗技术实时更新建模数据集,并根据预估精度决策在线优化和模型更新,提高工业软测量模型的在线估计精度。研究结果表明,基于滑动时间窗的LS SVM软测量建模方法,是一种有效的软测量建模方法。

Study and application of soft sensor modeling methods aiming at an industrial azeotropic distillation

The modeling method based on sliding time window was introduced,which dealed with these problems that the components of an azeotropic distillation were difficultly measured.An appropriate model was built by support vector machine technology based on least square.The data in sliding time window were used to optimize on-line or rebuild periodicity model to improve the estimate precision.The effective result indicates that the proposed method is a useful method for the modeling.