基于GIS信息系统及人工神经网络的砂岩富水性预测研究——以赵庄矿3号煤层顶板砂岩为例

在分析赵庄矿自然地理、区域构造、区域水文地质条件、矿井突水灾害的基础上,根据研究区的断裂构造、隔水层厚度、奥灰水压以及奥灰顶部富水性在空间区域上的变化特征,研发了基于GIS组件的顶板砂岩水预测信息系统,该系统主要实现了评价指标的自动统计、图形的管理等功能,具有强大的空间分析功能,同时应用非线性的人工神经网络(ANN)技术确定了各主控因素的权重系数,在此基础上建立了煤层顶板砂岩富水分区评价模型,大大提高了预测指标统计的速度和精度。并应用该系统对3号煤层顶板砂岩富水性进行了评价,为井田3号煤层顶板突水危险性评价与预测和井下水害防治提供理论依据。

3#coal seam roof and floor of sandstone water research in ZhaoZhuang coalmine

In analyzes the physical geography, regional geologic structure ,mine - hydrogeologic conditions and groudwater burst-ing characteristics in ZhaoZhuang coalmine, according to the research area＇ s faulted structure, aquiclude thickness, the Austria ash hydraulic pressure as well as Austria ash crown Fushui in the spatial region＇ s variation characteristic, has researched and de-veloped the GIS module＇ s roof sandstone water forecast information system （ PreSW 2010）.This system has mainly realized eval-uating indicator automatic：statistics, graph functions and so on,with the formidable spatial analysis function. Using the application misalignment artificial neural networks （ANN） technology determined each master control factor＇ s weighting factor, has based on this established coal bed roof sandstone Fushui district evaluation model, raising the forecast target statistics speed and the preci-sion greatly. And applied this system to carry on the No.3 coal bed roof sandstone. Those achievement provide theoretical basis for water inrush risk evaluation and forecast of floor of coalmine field No.3 and underground water inrush prevention.