Prediction of technical index for cobalt removal process based on SVM and chaotic PSO

Considering the characteristics of strong non-linearity and large time delay in cobalt removal process of zinc hydrometallurgy,a prediction method of technical index(cobalt concentration) combining the least square support vector machine(LS-SVM) and chaotic particle swarm optimization(CPSO) was proposed.CPSO used the mutation of the non-winner particles by chaotic search and mutation of the global best position by using small extent of disturbance to improve its search performance.The model parameters were optimized by CPSO and the input attributes were selected by binary PSO,which reduces the complexity and improves the prediction accuracy.The results show that prediction accuracy of the proposed model meets the technology requirements that the absolute error will be less than 0.1 mg/L when the solution concentration is less than 1 mg/L.