LS-SVR-Based Soft Sensor Model for Cement Clinker Calcination Process

A least squares support vector regression(LS-SVR) model for cement clinker calcination has been proposed, and successfully applied to an annual clinker production capacity of 0.73 million ton of Jiuganghongda Cement Plant in China. For the influence of unavoidable outliers in training sample on free calcium oxide (f-CaO) content and the degree of correlation between the original variables, a novel method based on Hampel identifier and principal component analysis (PCA) with outliers detection is discussed in detailed. In this method, outliers missing data points and deviating from normal values are detected. The PCA was applied to the model, which not only solved the linear correlation of the input variables, but also simplified the LS-SVR structure and improved the training speed. Industrial application results show that the soft sensor model has high accuracy and guidance to coal feeding to rotary kiln.