建立基于入库径流预报的长短期耦合梯级水库电力优化调度模型, 并以雅砻江锦屏一二级水电站为例, 探讨 了该电站入库径流量的预报效益。结果表明, 利用所建模型准确地获取锦屏梯级电站发电效益、预见期、预报精度 三者间的量化关系, 即预见期越长精度越高, 梯级电站平均年发电量越高, 据此提出锦屏梯级水库入库径流预报的 要求是预见期不应低于5 d, 预报精度平均相对绝对误差应小于01 15。

An o pt imal o per ation model o f hydro power stat ions in the cascaded r eser voir w as dev elo ped based on the coupling of long2term and shor t2term predictio n models of flow into the reserv oir . T he mo del was applied to the first2and second2g rade hy2 dro power st ations in the cascaded reserv oir of Jinping, and the pr edict ion benefit of flow int o the r eser vo ir was discussed. T he results showed that the mo del can pro vide the quantitativ e relationship bet ween pow er benefits, predictio n time, and predictio n accuracy o f the hy dropow er stations in the cascaded reservo ir of Jinping . In g ener al, lo ng er pr ediction time leads to higher pre2 dictio n accuracy and g reater pow er output s. It is sugg ested that the pr edict ion t ime o f flow int o the cascaded reserv oir of Jinping should not be less than 5 day s and the mean r elativ e abso lute err or o f predict ion sho uld be less than 0. 15.

中国水利水电科学研究院优秀青年科技人员科学研究专项( WE0145B072014) ; / 十二五0水专项/ 重点流域环境流量保障与容量 总量控制管理关键技术与应用示范0 课题( 2013ZX075012004)