Genetic algorithm based optimized least squares support vector machine for grate pressure prediction in grate coolers

As an essential part of cement clinker industrial production, the main task of the grate cooler is to cool the high temperature cement clinker; at the same time, the heat recovery of the high temperature gas generated during the cooling process. The working condition of the grate cooler and the cooling effect of the clinker is measured primarily by the pressure under the grate plate of the grate cooler. Therefore, a method based on genetic algorithm (GA) optimization of least squares support vector machine (LS_SVM) model parameters is proposed to predict the pressure under the grate of a grate cooler. The optimization calculation of LS_SVM model parameters is achieved through GA, which solves the limitation of LS_SVM prediction accuracy by model parameters.