非恒定耦合约束下的梯级水电多维协调优化调度

Large-scale hydropower cascade development changes the coupling relationship between hydropower stations, and uncertainty of natural runoff causes the drift of hydropower operation security domain. Meanwhile, such effects, in terms of generation benefits and wind power consumption, should be involved in the consideration of hydropower scheduling when integrating wind power. Against such background, several new challenges, in the aspects of intrinsic structure, external environment and achieved function, have been raised in hydropower system. Based on such challenges, the main work of this project involves: 1) Departure from the mechanism of water movement, complex couplings between the cascade hydropower stations are described in time and quantity dimensions, and the cascade hydropower scheduling model by considering the unsteady coupling constraints is established. 2) Combined with the impacts of the correlation between basin cascade hydropower stations, spatiotemporal metastatic of runoff and extreme weather, the explicit expression of hydro random security domain is studied, and the cascade hydropower robust optimal scheduling method with the hydro random security domain is proposed. 3) For the stochastic volatility of wind power, wind reserve capacity demands in different time scales are quantified by decoupling method. Accordingly, the coordination allocation strategy of peaking and frequency regulating capacity of hydropower stations is drafted. In order to completely utilize the storage and regulating abilities of hydropower, the multi-scale hydropower coordination compensation dispatching mode is proposed, which can satisfy the peaking and frequency regulating demand of wind. The achievement of this project can provide theoretical and methodological support for hydropower multidimensional optimal scheduling.

大规模水电的梯级开发改变了水站之间的水耦合关系，天然来水的不确定性造成了水电系统运行的安全域的漂移，风电的并网使得水电调度需要兼顾自身效益和风电的消纳。针对上述水电系统在本征结构、外部环境以及实现功能面临的新挑战，项目主要开展了以下3方面的研究：1）从水流运动过程的机理出发，沿时间和空间两个维度对梯级水电站之间的复杂耦合关系进行描述，建立考虑非恒定耦合约束的梯级水电优化调度模型；2）结合梯级水电站之间的流域相关性、来水的时空转移性研究水电的随机安全域，提出基于随机安全域的梯级水电鲁棒优化调度方法；3）针对风电的随机波动性，通过解耦的方法量化不同时间尺度上风电备用的需求，进而制定水电调峰调频容量协调分配策略，充分发挥水电的蓄能和调节能力，构建面向风电并网调峰调频需求的水电多尺度协调补偿调度模式。本项目的研究成果将为水电站群多维优化调度提供理论和方法支持。

近年来，我国流域梯级水电呈现规模化发展，形成了一批大规模的梯级水电站群。在水电优化调度问题中面临水电站群间水力、电力的复杂联系、流量演进过程建模、不确定性等技术难点，因此开展了非恒定耦合约束下的梯级水电多维协调优化调度研究，主要内容如下：.首先，针对考虑非恒定耦合约束的梯级水电优化调度问题，考虑到上下级水电站间耦合关系的变化，分别结合非汛期和汛期两个调度场景进行了研究。在非汛期，建立了一种新的具有连续变化的水流延时的短期调度模型，并提出了逐次逼近方法予以求解。在汛期，考虑到梯级水电流域来水充沛，运用马斯京根法建立了上下游流量的耦合关系，并利用混合整数线性规划法予以求解。算例分析表明，考虑非恒定耦合关系的梯级水电优化调度模型精确性更高。.然后，针对基于随机安全域的梯级水电鲁棒优化调度问题，为保证所有可能来水情况下系统均能够安全运行，引入了鲁棒优化理论。同时考虑到梯级水电站之间来水的时序分布特性以及流域相关性，提出了多区间不确定集来描述水电运行的随机因素，克服了传统鲁棒优化的保守性。进而构建了经济性和安全性协调的两阶段鲁棒优化模型，并结合对偶理论及混合整数二次规划法进行了求解。算例分析表明，所提的鲁棒优化调度方法比场景规划法求解效率更高，并且多区间不确定集可以有效地改善调度方案的经济性。.最后，针对面向风电并网调峰调频需求的水电协调补偿调度问题，为减少风电出力、负荷和备用价格不确定性对系统运行的影响，运用离散傅里叶变换法量化出了风电和负荷预测误差引起的多种类型备用的需求，进而考虑备用价格的不确定性以及风电、梯级水电和火电的多种运行约束，构建了风险规避型的多能源协调优化调度模型，然后将其转换为混合整数线性规划模型进行求解，并在求解过程中应用含简森不等式约束和多割集反馈两种优化策略的Benders分解法进行求解，在确保计算精度的同时提高求解效率。最后通过算例分析验证了所提方法和模型的有效性。