Application of Soft Sensor Based on LS-SVM on Estimation of Alumina Powder Flow

Alumina powder flow in electrolytic aluminum plant for the production of alumina can not be precise measured online, the Least Squares Support Vector Machines (LS -SVM) was applied in the modeling of alumina powder flow estimation in the process of alumina conveying in this paper, and the soft sensor model based on LS - SVM was compared with the soft sensor model based on RBF. The result of simulation research proves that the soft sensor model based on LS-SVM has a higher precision accuracy and better generalization ability. The soft sensor technique is effective in estimate accuracy of the alumina powder flow.