数据挖掘技术在辽河水文预报中的研究

Study of used up data technique in hydrological forecast of Liao river

针对以数据仓库为基础预报的特点,结合我国的防洪减灾实际,建立面向预报主题的水文数据仓库的结构模型,并对其结构、功能、数据存储模型和实现技术进行研究,从而对海量水文信息进行管理与分析,建立包括数据层、组织层、挖掘层和决策层的水文数据仓库体系,各层承担着水文数据挖掘不同阶段的任务,从数据预处理、数据挖掘到知识表达,形成了一个完整的系统体系。本文结合数据仓库中的聚类、分类、关联分析等方法特点,着重探讨了在水文预报中的应用。

According to the characteristics of forecast based on data warehouse,combing with the reality of flood control and disaster reduction,the paper builds the hydrological data warehouse model oriented to the forecast for the analysis and management of massive data,studies the structure,function,data storage model and implementation technology of the model,builds a hydrological data warehouse system with data layer,organizing layer,used up layer and decision layer which undertakes different work in data use.And a complete system is formed from data preprocessing,data used up to knowledge representation.The paper discusses the application in hydrological forecast combined with the characteristics of methods such as cluster,classification and association analysis.