基于模糊推理和神经网络的采矿方法智能决策系统  

INTELLIGENT DECISION SUPPORT SYSTEM OF MINING METHOD BASED ON FUZZY EXPERT SYSTEM AND NEURAL NET

作　　者：黄德镛[1] 胡运权[1] 陈孝华[2] 杨德全 叶加冕[2] 宦秉炼[2] 张庆文 

机构地区：[1]哈尔滨工业大学,哈尔滨150001 [2]昆明理工大学,昆明650093 [3]云南锡业公司,云南个旧661400

出　　处：《有色金属》2002年第2期104-107,共4页Nonferrous Metals

基　　金：云南省国际科技合作项目 (中国 -葡萄牙 )(编号 :97D0 0 7)

摘　　要：有针对性地引入人工智能、专家系统、人工神经网络和计算机辅助设计等先进技术 ,通过有效的组合与集成 ,形成功能先进的智能化决策支持系统。采矿方法智能决策系统由四部分组成 :将矿体的赋存条件转化为模糊知识和模糊规则 ;运用模糊推理专家系统初选可行方案 ;用人工神经网络对可行方案的技术经济指标进行预测 ;通过群体多目标决策 ,优选出实施方案。应用于云南锡业公司松树脚矿 ,用优选方案进行工艺试验 ,取得了贫化损失小。The intelligent decision support system of mining method based on fuzzy expert system and neural net is established by combination of artificial intelligence, expert system, neural net, and computer assistant design system. MMIDSS includs four parts. In the first part, the deposit character is denoted by fuzzy knowledge and fuzzy rules; in the second, feasible approaches are selected by the fuzzy expert system; some primary technic and economic indexes are estimated by ANN in the third; the optimum implement scheme is obtained by multi objective group decision in the last part. MMIDSS is used in the Songshujiao Mine in Yunnan Tin Company, the results of experiment with the optimum method determined by MMIDSS show that the dilution and production cost are low, so the system design is reasonable, valuable and applicable.

关 键 词：模糊推理 神经网络 采矿方法 智能决策系统 

分 类 号：TD853.3[矿业工程—金属矿开采] TP183[矿业工程—矿山开采]