水利学报Issue(10)：1189-1196,1203,9.
基于降雨预报信息的梯级水电站不确定优化调度研究Ⅱ：耦合短、中期预报信息
Stochastic optimization operation for cascade hydropower reservoirs by using precipitation forecastsⅡ. Coupling short and medium-term information
摘要
Abstract
This paper presents the Two Stage Bayesian Stochastic Dynamic Programming (TS-BSDP) model to use the forecasting inflow from Quantitative Precipitation Forecasts of Global Forecast System (QPF-GFS). Firstly, this model divides the forecasting inflow (e.g., 10 days) into two parts. The near part (e.g., 1~5 days) is assumed as accurate one. While the rest (e.g., 6~10 days) needs to account the uncertainty, which addressed by Bayesian Decision Theory. Secondly, the TS-BSDP model runs iteratinely to derive the hydropower operation policies, which used to make operation decision for near part by rolling forward or the entire forecasting horizon. In this paper, China’s Hun River cascade hydropower reservoirs system is taken as an example, and the values of QPF-GFS (10 days lead time) are used to forecast the 10 days inflow. The operation policies for 1~5 days and entire 10 days incorporate with forecasting inflow to simulate the decision processes by 5 days rolling forward and entire 10 days, respectively. Finally, the simulation results demonstrate that the performance of rolling forward by 5 days is better than the other de-cision processes.关键词
水库群/数值降雨预报/预报信息套接/聚合分解/贝叶斯随机动态规划Key words
bayesian stochastic dynamic programming/quantitative precipitation forecasts/inflow forecast-ing/hydropower operation/decision processes分类
建筑与水利引用本文复制引用
徐炜,彭勇,张弛,王本德..基于降雨预报信息的梯级水电站不确定优化调度研究Ⅱ：耦合短、中期预报信息[J].水利学报,2013,(10):1189-1196,1203,9.基金项目
国家自然科学基金资助项目（51109025）；教育部博士点基金（20100041120004）；中央高校基本科研业务费专项（DUT13JS06）；水利部公益性行业专项 （）
