Soft Sensing Based on LS-SVM and Its Application to a Distillation Column

Dry point of aviation kerosene in the atmospheric distillation column is a very important process value for quality controlling. But unfortunately few on-line hardware sensors are available to this value or such sensors are difficult to maintain. This paper adopts a novel method based on least squares support vector machine (LS-SVM) regression to implement on-line estimation of aviation kerosene dry point. Compared to traditional radial basis function (RBF) neural network and squares support vector machine (SVM) regression methods, using the same sample data, the simulation results show that the soft sensing based on LS-SVM regression has better abilities of model generalization and real-time character