The research of least squares support vector machine optimized by particle swarm optimization algorithm in the simulation MBR prediction

Abstract

This paper proposes an intelligent algorithm to predict the MBR membrane flux. The algorithm applies the least squares support vector machine (LS-SVM) to the research of MBR simulation prediction, optimize the penalty factor and kernel parameters of LS-SVM model by particle swarm optimization (PSO) for avoiding the blindness of artificial selection parameter. Due to the complexity and cross-cutting of the factors that affect MBR membrane fouling, first of all, we analyze the factors by principal component analysis (PCA), extract the important factors as the LS-SVM input layer, MBR membrane flux as output layer, and then create PSO-LSSVM prediction simulation model. In the end, we get predictive results with the model. By comparing the predicted results with experimental data, the algorithm has higher prediction accuracy for MBR membrane flux. To further verify the effectiveness of the algorithm, we also compare the model with BP neural network model, the results show that the prediction model of PSO-LSSVM has a higher prediction accuracy.