炉渣含金量预测的神经网络模型

NEURAL NETWORK MODEL FOR GOLD CONTENT ESTIMATION IN SLAG

金泥的处理工业上普遍采用火法冶金，炉渣组成对获取最高的金回收，率和最低渣中含金量具有重要影响，本文应用神经网络方法，对苏打－硼砂－玻璃－食盐所组成的渣系同熔炼渣含金量的关系进行研究，建立可用于预测不同炉渣组成分渣含金的神经网络模型，包括网络类型，网络结构及其算法），研究结果表明，所建立的三层反向传播神经网络模型可以用于熔炼体系渣含金的预测，且比传统回归分析法有许多突出的优点。

Pyrometallurgy is often used in the industrial process for treating the gold slime.Slagcompositions have an important influence on maximum gold recovery and on theminimum gold content in the slag.The relationship between slag compositions based on soda borax silica glass salt system and gold content in slag was investigated by using neural network in this pater.The neural network model for estimatiog the gold content of different slag compositions was presented,including the neural network type,structure and its learning algorithms.The application study indicates that the three layer back propagation neural network model can be applicd to estimate gold content in slag.Compared with traditional regression method,neural method has many advantages.