基于秩相关随机变量模拟的水库防洪风险估计

Risk Evaluation Model of Reservoir Flood Control Based on Sampling Rank Correlated Random Variables

采用基于Spearman秩相关随机变量模拟方法,将拉丁抽样法产生的独立洪峰洪量序列转换成满足特定相关结构的洪峰洪量序列,利用模拟的洪峰和时段洪量放大修匀典型过程线生成入库洪水过程,实现了入库洪水过程随机模拟。采用Monte Carlo法估计防洪风险,构建了水库防洪风险估计模型。潘家口水库汛限水位调整防洪风险估计应用结果表明:主汛期汛限水位上限值控制在218.00 m,所建立的入库洪水过程模拟方法简便有效,非常适合于复杂洪水过程模拟,构建的防洪风险分析模型计算结果客观合理,在洪水资源安全利用系统风险分析中具有推广应用价值。

A flood risk evaluation model based on sampling rank correlated random variables is introduced.In the model the Latin sampling method is used to generate independent flood peak and volume series,and a correla-ted random variables sampling method inducing the desired spearman correlation rank correlation coefficient matrix is used to convert the independent series to related series.The flood process is amplified and smoothed by the typi-cal hydrograph and the simulated flood peak and volume series.The flood control risk rate is evaluated by Monte Carlo simulation.This model is applied to Panjiakou reservoir located in Luanhe river basin.Two conclusions are reached：1）The upper limit of flood control water level for the main flood season is 218.00 m;2）The flood process simulation model is simple to use and effective and the flood risk analysis result was objective and reasona-ble.The proposed method is suitable for risk analysis of flood resources utilization.