Solving the Transient Cost-Related Optimization Problem for Copper Flash Smelting Process with Legendre Pseudospectral Method

Abstract

In the present contribution, a scheme to solve transient cost-related optimization problem of dynamic transient procedure for copper flash smelting process is investigated. Taking the actual copper flash smelting process at a Smelter in China as the research object, the transient costrelated optimization problem is presented by integrating transient time, transient resources consumption and the state fluctuation constraint. Then, it was considered as the operational parameters trajectories optimization problem that generates minimum-time and minimum-resources consumption during dynamic transient procedure. The dynamic relationship and the working state fluctuation are constructed in the form of constraints in the resulting optimization problem formulation. The Legendre Pseudospectral method is used to solve the constrained, nonlinear optimization problem. The results of numerical experiments of dynamic transient procedure for copper flash smelting process are given to illustrate the proposed scheme.

Statistical modeling of copper losses in the silicate slag of the sulfide concentrate smelting process

“…R 2 , which is the coeffi cient of determination, is the squared value of the multiple correlation coeffi cient. It shows that approximately 19 pct of variation in Y is explained by the model, as is indicated already by the regression to residual ratio 17 . Accordingly, it can be concluded that the MLRA approach would not be the appropriate tool for modeling of the investigated process, because of low values of all adequacy tests indicators.…”

“…However, the ratio of regression to residual is 19 pct: 81 pct, advocating that only 19 pct of the dependent variable (Y) values are explained by the model. Where ''Regression'' displays information about the variation accounted for by the model, and ''residual'' displays information about the variation that is not accounted for by the model 17 . This is a clear indicator that the MLRA model of investigated data set would not result in adequate enough accuracy.…”