数据仓库和数据挖掘技术在滑坡预测预报中的应用

Application of Data Warehouse and Data Mining on Landslide Prediction

为从滑坡历史数据中获取有效的预测预报信息,利用基于数据仓库的数据挖掘技术,结合已有滑坡多个维度的数据集,以巴东县新城区滑坡灾害分布作为研究对象,建立了滑坡灾害敏感性和灾害区划模型。结果表明:该模型对于滑坡空间分布的预测精度可达到87.5%左右,对于滑坡时间尺度上的预测精度较低,约为65%左右,其准确度可以满足工程要求;数据仓库和数据挖掘技术在地质灾害预测预报领域具有广泛的应用前景,较传统预测预报方法更为方便和迅捷。

In order to obtain effective prediction from the historical landslide data,the data mining technology based on data warehouse were used.Combined with the existing landslide multi-dimensional data set,taking the landslide disaster distribution in the new urban area of Badong County as the research object,the landslide disaster sensitivity and disaster zoning model are established.The result shows that the prediction accuracy of the model for the spatial distribution of landslide can reach about 87.5%,and the prediction accuracy on the time scale of landslide is low,about 65%.Its accuracy can meet the engineering requirements.The data warehouse and data mining technology in the field of geological disaster forecast has wide application prospect,which is more convenient and rapid than traditional prediction methods.