Prediction for Matte Grade in the Process of Copper Flash Smelting Based on QPSO-LSSVM

According to the complexity of the reaction mechanism and the requirement of the craft indicator during the process of copper flash smelting, the prediction model of matte grade was proposed by combining quantum particle swarm optimization algorithm (QPSO) with least squares support vector machine (LS-SVM) in this paper. Firstly, the nonlinear relation model between matte grade and craft indicators in copper flash smelting process was established by using the LS-SVM. Secondly, the parameters of LS-SVM were optimized by using the QPSO algorithm. Finally, the simulation results show that the maximum relative error of the matte grade is 0.47% and the relative root mean square error is 0.33%.Results indicate that the model can satisfy the requirement of production process and can be used to guide the practical production.