采矿方法模糊优选的神经网络模型构造及其应用  

Neural Network Model Construction and Its Application in Fuzzily Optimization of Mining Method

作　　者：吴爱祥[1] 郭立[1] 余健[1] 阳雨平[1] 肖雄[1] 

机构地区：[1]中南大学资源安全与工程学院,湖南长沙410083

出　　处：《矿冶工程》2003年第3期6-8,11,共4页Mining and Metallurgical Engineering

基　　金：国家自然科学基金资助项目 (50 0 740 34);湖南省自然科学基金项目 (0 2JJY2 0 78);高等学校优秀青年教师教学科研奖励计划资助

摘　　要：应用神经网络模式识别原理 ,构建一种人工神经网络 (ANN)———弹性神经网络 ,能较好地解决常用BP神经网络梯度算法中可能出现的收敛缓慢和局部最小问题 ,将模型应用于大红山铜矿缓倾斜中厚矿体采矿方法模糊优选 ,与用模糊数学法得出的结论基本一致。终选出的采矿方法工业试验后取得了显著成效 。A ANN(artificial neural network)--an elastic neural network is constructed by applying the identifying principle of neural network patterns, overcoming better the show convergence and local minimium occurring in BP neural network grads algorithm. The ANN has been applied to fuzzy optimization of mining methods of show obliquity and middle thickness orebody in Dahongshan Copper mine. The results obtained are about the same as the results obtained using fuzzy mathematics. Great success has been achieved using the finally selected mining methods in the industrial tests. The ANN has proved to be feasible.

关 键 词：采矿方法 弹性神经网络 模糊优选 递推算法 模式识别 缓倾斜中厚矿体 

分 类 号：TD853.211[矿业工程—金属矿开采] TP183[矿业工程—矿山开采]