论文部分内容阅读
针对大型梯级水电年末预留水位优化问题，提出在给定置信度条件下的来水在险值(inflow at risk，IaR)方法，用以计算第1年可靠来水流量，并构建以发电收益最大化为目标的全景梯级水电跨年随机优化调度模型，使得年末预留水位能够适应第2年多种来水情况下的运行约束。采用条件风险价值(conditional value at risk，CVaR)指标衡量发电收益风险，并建立将发电收益经济性与风险性统一的双层规划模型。针对所建混合整数线性规划模型的解算，提出了主、子问题一体控制的最优奔德斯(Benders)分解方法，并对模型进行求解。对一个2级梯级水电站系统进行仿真分析，结果表明，所提出的全景跨年随机优化调度模型在不同风险偏好下，可以有效提高梯级水电站的跨年发电收益、减少弃水量，同时验证了最优奔德斯分解策略的高效性。“,”As for the reservation optimization of the end-year water level for the large cascade hydropower plant, the inflow at risk (IaR) method under the given confidence level was proposed to calculate the dependable inflow of the first year, and a full-scenario biennial stochastic scheduling model of cascade hydropower plant was proposed to maximize the generation profits which will adapt the optimized water level to the constraints under different scenarios in the following year. The conditional value at risk (CVaR) was applied to illustrate the risk of the generation profit, and the bi-level programming combing the economy and the risk of the generation profit was constructed. As for the solution of the proposed mixed-integer linear programming model, the optimal Benders decomposition with the integrated control of the master problem and the subproblems was proposed to solve the model. A 2-cascaded station was utilized to test the proposed model, and the results show that the total generation profit is improved, the average spillage was reduced, and the high efficiency of the optimal Benders decomposition strategy was also proved.