人工神经网络在高梯度磁选试验研究中的应用  

MODELLING OF HIGH GRADIENT MAGNETIC SEPARATION WITH ARTIFICIAL NEURAL NETS

作　　者：向发柱[1] 何平波[1] 陈荩[1] 

机构地区：[1]中南工业大学矿物工程系

出　　处：《矿产保护与利用》1998年第2期22-26,共5页Conservation and Utilization of Mineral Resources

摘　　要：用人工神经网络建立了高梯度磁选过程模型，对不同隐含层节点数的神经网络模型预测性能进行了评价。并对高梯度磁选过程进行了模拟研究，结果表明，在相当宽的操作范围内，模型有比较好的预测结果，这说明建立的模型是合理的、可行的。A sigmoidal backpropagation neural network model for high gradient magnetic separation was developed in this paper. The performances of various network models with different numbers of hidden nodes were evaluated. The application of the sigmoidal backpropagation neural network model to the simulation of high gradient magnetic separation processes of chalcopyrite was illustrated. The simulation results show that the model is capable of making a good prediction over a broad range of operating conditions and the developed model is reasonable and feasible.

关 键 词：人工神经网络 高梯度磁选 试验研究 选矿 

分 类 号：TD924.1[矿业工程—选矿]