BP自组织神经网络在地下水动态分类中的应用

BP neural network applied in the groundwater regime classification

本文是根据国内外较为流行的BP自组织神经网络方法,与区域地下水动态成因及地下水观测井历时曲线形态相结合对哈尔滨市地下水的动态型进行定量分类:波动—水文型、上升—弱水文型、上升·下降—开采型、上升—开采型,并且定位在图上。此种分类更加直观地反映本区地下水在空间和时间的变化规律,并对本区地下水的水流模型参数分区具有参考价值。

We classified the types of groundwater regime only qualitative. In this paper, the BP neural network is applied in the groundwater regime, in considering hydrograph of groundwater level and groundwater level contour map. The types of groundwater regime in Harbin city are classified into four kinds in quantitative: wavehydrology, upweak hydrology, up·downwithdrawal, upwithdrawal. It can show the groundwater spatiotemporal exchange.