Modeling for cement combined grinding system with closed grinding process based on LS-SVM in typical working condition

Abstract

In order to overcome the difficulty in the modeling of combined grinding system, an identification algorithm based on the least squares support vector machine (LS-SVM) is applied to the modeling of the closed grinding process in typical working condition as a new tool. The model parameters are selected by analysis of its process mechanism. The sample data sets are obtained from the historical data of on-site production in typical working condition. As the LS-SVM algorithm is of the advantages of simple structure and high speed, according to the mechanisms of LS-SVM algorithms, a model describing the variation of out-mill elevator current was obtained by modeling simulation. Compared with the model based on artificial neural networks (ANN) method, the proposed model has better capability of generalization and better convergent speed if a small data set is available. The simulation results show that the LS-SVM is more valuable when applied to the modeling of the cement combined grinding process.