Soft sensing for dry point of the aviation kerosene based on PLS and LS-SVM method

In the distillation process, many important process variables are often difficult to be measured online. For example, the aviation kerosene is an important index of operation quality, but current methods cannot obtain the real-time value of the aviation kerosene efficiently. To solve this problem, a method of selecting the input variable based on partial least squares regression (PLS) is proposed. This method chooses the cumulative variance contribution rate as the standard, and the factors that affect the cumulative explanatory ability of the principal elements will be discussed. The optimal subset will be selected from 104 variables by using the partial least squares regression, then the optimal combination will be determined to improve the accuracy of the model. The least squares support vector machine (LS-SVM) regression is selected to establish the soft sensor model of the aviation kerosene. The simulation results show that this method is fast and can improve the prediction accuracy of aviation kerosene and reduce the computational complexity of the model.