点击次数：

论文类型：期刊论文

发表时间：2011-11-01

发表刊物：ICIC Express Letters

收录刊物：EI、Scopus

卷号：5

期号：11

页面范围：4227-4232

ISSN号：1881803X

摘要：Aviation kerosene is one of the products produced by the atmospheric distillation column. The dry point is an important quality parameter to aviation kerosene, but the hardware sensor is not available for this parameter. An inferential sensor which adopts the kernel ridge regression (KRR) method to implement on-line estimation of aviation kerosene dry point is proposed. Through collecting data from the industrial distillation process and building the KRR model, the simulation results show that KRR provides a new and effective method for soft sensing modelling and that the inferential sensor has better abilities of model generalization. ? 2011 ICIC International.