人工神经网络在矿井突水预报中的应用

突水预报是一项重要的矿井水文地质工作。借助于人工神经网络在处理非线性问题或非结构问题方面的优势,采用BP算法,基于大量矿井突水样本实例建立了突水预报神经网络模型,并将该模型用于实际预报,并取得了较好的效果。结果表明,模型具有较强的实用性。为了提高模型的预测精度,在训练样本的选择上还应具有一定的代表性。

The application of artificial neural network theory to mine water inrush prediction

Water inrush prediction is an important mine hydrogeological work. With the advantage of artificial neural network in solving nonlinear or unstructured problems and by using BP algorithm, a neural network model for predicting mine water inrush has been built based on a large number of water inrush examples. To apply this model to practical prediction produces a better effectiveness, which shows that this model can be practically used. In order to improve the accuracy of model prediction, the training samples should be representatively prepared.