铜闪速熔炼影响规律的神经网络分析  

Analysis of the Effect Factors of Copper Flash Smelting Based on Neural Network

作　　者：汪金良[1] 卢宏[2] 汪仁良 曾青云[1] 

机构地区：[1]江西理工大学材料与化学工程学院,赣州341000 [2]江西理工大学信息工程学院,赣州341000 [3]贵溪冶炼厂,江西贵溪335424

出　　处：《有色金属（冶炼部分）》2008年第2期2-5,共4页Nonferrous Metals（Extractive Metallurgy)

基　　金：国家自然科学基金项目(50364004);江西省自然科学基金项目(0250026)

摘　　要：基于已建立的神经网络模型,研究了富化率、吨矿氧量、熔剂率以及铜精矿主要成分对铜闪速熔炼过程的影响。结果表明:富化率的增大会使铜锍品位降低、铜锍温度升高,而对渣含Fe/SiO2影响不大;吨矿氧量的增加会使铜锍品位、铜锍温度及渣含Fe/SiO2都升高;熔剂率的增加会使渣含Fe/SiO2明显下降;精矿中Cu含量的增大会使铜锍品位升高,铜锍温度稍微降低;而Fe的影响与Cu相反;S/Cu一般控制在1.0±0.2,自热熔炼应控制在1.34以上。The effects of the oxygen grade, the oxygen volume per ton concentrate, the flux rate and the elements content in copper concentrate on the copper flash smelting process are studied based on the built neural network model. Results show that the matte grade reduces, the matte temperature increases but the Fe/SiO2 in slag changes little when the oxygen grade increases; the matte grade, the matte temperature and the Fe/SiO2 in slag increase all when the oxygen volume per ton concentrate increases; the Fe/SiO2 in slag drops clearly when the flux ratio increases; the increment of the Cu content in concentrate makes the matte grade increase but the matte temperature drop little; the effect of the Fe content in concentrate is opposite to the Cu content; S/Cu should be controlled from 0.8 to 1.2 generally, but more than 1.34 for self-heat smelting.

关 键 词：神经网络 闪速熔炼 铜 因素 

分 类 号：TF811[冶金工程—有色金属冶金]