摘要
受暴雨的天气系统、降雨强度、降雨中心时空分布、下垫面特性和人类活动等因子的影响,洪水过程形成具有高度的复杂性和不确定性,同时又表现出很强的规律性。针对这一现象,本文应用二阶聚类方法和粗糙集理论建立实时洪水分类预报模型。该模型基于影响因子集利用二阶聚类方法对历史洪水进行聚类,并深入的分析各因子对洪水分类的影响程度;采用遗传算法率定相应洪水类型的模型参数,并通过粗糙集挖掘影响因子与洪水类型间的隐含关系;在实际应用中,根据当前获得的洪水信息识别出所发生洪水的类型并选择相应的模型参数进行洪水预报。将所建立的模型应用于观音阁水库洪水预报方案研究中,实验结果表明,该方法能够准确、迅速的判断洪水类型并选择相应预报模型参数,能有效提高水库实时洪水预报精度。
This paper presents a new classified real-time flood forecasting framework that integrates two step cluster and rough set with a conceptual hydrological model.The two step cluster was used to classify historical floods based on flood antecedent impact factors and time-varying rainfall information,the conceptual model was calibrated for each type of flood,and the rough set theory was used for mining the relationship between flood information and flood types.In application,flood types are identified from the flood information obtained,and the related model parameters are selected for the forecasting.Application to Guanyingge reservoir in Liaoning province shows that this framework can identify the type of flood and forecast flood fast and accurately,improving the real-time forecasting precision significantly.
出处
《水力发电学报》
EI
CSCD
北大核心
2013年第2期60-67,共8页
Journal of Hydroelectric Engineering
基金
国家自然科学基金(批准号:50809011
50909012)
关键词
水文学
洪水分类
实时预报
粗糙集
二阶聚类
hydrology
flood classification
real-time forecast
rough set
two step cluster
