Prediction Model of Operating Parameters of Copper Flash Smelting
Received:November 25, 2014  Revised:November 26, 2014
View Full Text  View/Add Comment  Download reader
DOI:10.3969/j.issn.1007-7545.2015.05.002
KeyWord:copper flash smelting; on-line control; back propagation neural network; prediction model
           
AuthorInstitution
XIE Kai 中南大学 能源科学与工程学院
misha 中南大学 能源科学与工程学院
YAN Bing 中南大学
LI Qi 中南大学
Hits: 387
Download times: 1029
Abstract:
      Based on practical copper flash smelting and control process, neural network model was studied in copper flash smelting process on-line control. Based on analysis of the effecting factors on solvent rate, oxygen consumption of melting, and air volume of reaction tower, a back propagation neural network prediction model was presented to predict the above parameters. BP neural network models with input vector containing only main elements and impurity elements in concern were established. The simulative results show that the maximum relative error between output value and actual value is less than 1.0%. Output value accords with practical data very well. The mode including impurity elements in input parameters has higher calculation accuracy.
Close