人工神经网络技术在水文地质学中的应用潜力初探

The application potentiality of ANN technique on hydrogeoloy

在许多水文地质问题中，多因素且非线性的影响常使传统的集中参数随机模型或分布参数确定性数值模型的方法难以对其作出符合实际的评价与预测。本文从几个典型的水文地质问题入手，利用人工神经网络技术的高度自组织、自适应与自学习能力和分类计算能力，对这些问题的解决进行了系统的ＢＰ网分析。结果表明，人工神经网络的应用可有效减少人为的主观臆断性，其训练识别的结果更符合实际，效果令人满意，因此具有十分广阔的应用前景。

In many hydrogeological problems,especially in some nonlinear ones,it is difficult for the traditional methods of random concentrated parameter model and determined distributed parameter model to evaluate and predict actual cases.On the basis of analyses of several typical hydrogeological problems,this paper deals with some cases using the artificial neural network(ANN) technique with functions of high ability of self organization,self adapting,self training and classified calculation.It is obvious that the application of ANN can reduce the human's subjective erroes efficiently by means of systematic analysis of BP network,and the results are much better than the traditional methods.Therefore,it should be said that the ANN technique has a bright application future on hydrogeology and engineering geology.