空间变尺度流域洪水预报修正研究

Research on Spatial Scaleable Flow Forecasting Correction for Certain Drainage Areas

针对考虑流域内水库运行规律但缺少小型水库、塘坝运行资料的洪水预报,提出了一种以朴素贝叶斯分类为基础、利用小尺度流域水库信息推求大尺度全流域水库信息的空间变尺度水库运行规律分析方法,并将其分析结果应用于流域洪水预报修正中。最后,以丰满五道沟控制站以上流域为例,采用该方法对2场洪水预报进行了修正。结果表明,修正后的预报精度较原预报方案预报精度明显提高,预报结果可以对水库调度起指导作用。

As certain flowforecasting takes the reservoirs＇ running disciplinarian into account but lacks correlative small-scale reservoir and embankment running data for certain drainage area,the analytical method is introduced based on naive Bayes classification for spatial scaleable reservoirs＇ running disciplinarian,which can forecast big scaleable reservoir data for the full drainage area by using correlative data of small scaleable reservoir and apply the analytical results to flow forecasting.Then,the flow forecasting of two flood disasters are corrected by the established method,withthe upriver drainage area beyond Fengman＇s Wudaogou Control Station taken as an example.The results showthat the precision of correctional flowforecasting is improved dramatically and the flood forecasting correctional results can play a leading role in the operation of reservoirs.