基于神经网络的放出体形态研究
APPLICATION OF FUNCTION-LINKAGE NEURAL NETWORK TO ANALYSIS OF SHAPE OF ORE-ROCK MASS DRAWN DOWN
-
摘要: 采用崩落法采矿时,放出体受多种具有非线性关系因素影响而呈现各种不同的形态,很难用与实际相一致的数学模型加以描述。利用神经网络求解非线性问题的优点,建立了函数联结神经网络模型。通过对实验观测数据进行学习和训练,可以实现对放出体形态的预测,其结果和实际观测值基本一致。结果表明该方法预测精度高,实用可行。Abstract: In the case of caving mining method,the ore rock mass drawn down has various shapes due to many influencing factors.It is difficult to predict exactly these shapes by means of a common mathematic model.A function linkage neutral network is introduced that is able to predict exactly the shape of ore rock mass drawn down,thus providing the possibility to reduce ore depletion.The predicted shape of ore rock mass conforms basically with the result from test.Besides,the network is also highly reliable and easily practicable.
下载: