基于人工神经网络的选矿指标预报专家系统  

AN ANN-BASED EXPERT SYSTEM FOR INDEX FORECAST OF ORE DRESSING

作　　者：李克庆[1] 闫飞[1] 周伟 

机构地区：[1]北京科技大学,北京100083 [2]鞍钢弓长岭矿业公司,辽阳111007

出　　处：《中国矿业》2004年第8期56-58,共3页China Mining Magazine

摘　　要：针对国内多数铁矿山矿石来源复杂、选矿技术指标难于准确预测的问题 ,本文提出通过建立人工神经网络模型 ,构筑选矿指标预报专家系统 ,进而优化供矿结构的方法。通过编制相应的计算机程序 ,用实例检验证明了模型和方法的可靠性和实用性。According to the ore source complexity of many iron mines and the difficultance for accurate forecast of ore dressing index,this paper puts forward a new method to optimize ore mixing structure through establishing artificial nerve network model and building the index forecast expert system of ore dressing.The reliability and feasibility of the model and method are testified by instances.

关 键 词：配矿 神经网络 选矿指标 

分 类 号：N945.1[自然科学总论—系统科学]