基于改进灰关联分析的相似洪水动态识别与预报方法

实时洪水预报是优化流域防洪减灾方案的重要支撑信息,基于洪水相似性的相似洪水预报是一种延长实时洪水预报预见期的可行途径。为改进灰关联分析法在相似洪水识别与预报中的应用效果,通过引入置信度距离、离势系数权重,加强传统灰关联方法的抗差能力,减少相似性分析过程中的信息损失,并将改进后的方法应用于相似洪水识别与预报过程中,最后结合池潭水库实测雨洪资料进行验证分析,预报精度评价结果表明,该方法预报精度优于传统灰关联分析法,可作为实时洪水预报的辅助方法。

Dynamic Similar Flood Recognition and Forecasting Method Based on Improved Gray Correlation Analysis

Real-time flood forecasting is important supporting information in optimizing flood control and disaster reduction scheme. Similar flood forecasting can be utilized to extend lead time of real-time flood forecasting. In order to improve the application of gray correlation analysis in similar flood recognition and forecasting, we introduced the confidence distance and the coefficient of variation to enhance robustness and reduce information loss. Aforementioned method was applied in the dynamic similar flood recognition and forecasting in Chitan reservoir, the precision assessment results indicate that the forecasting precision of this method is better than that of the traditional gray correlation analysis method, and it can be used as an auxiliary means for real-time flood forecasting.