基于数据驱动分布鲁棒优化的梯级水光蓄联合优化调度

具有不确定性复杂耦合的多种可再生能源的不断渗透给当前电力系统协同调度带来了巨大挑战。传统的随机优化(SO)及鲁棒优化(RO)方法由于难以获得精确概率分布函数及优化结果过于保守使其应用大大受限。本文基于数据驱动分布鲁棒优化理论(data-driven DRO),提出了梯级水光蓄联合发电系统协同调度方法。该方法首先考虑系统互补经济调度成本建立两阶段调度模型,制定各电站出力调度计划,然后引入综合范数约束限定概率置信区间,并考虑最恶劣分布下的实时运行调整成本,获取日前调度计划的最优调整方案。两阶段协同调度模型采用MP-SP框架,引入CCG算法展开求解,日调度计划和调度调整方案形成最优调度计划。引入示范区实际运行数据开展实例验证,所提互补联合调度方法的有效性及高效性得以验证。

Data-driven Distributionally Robust Optimization Based Coordinated Dispatching for Cascaded Hydro-PV-PSH Combined System

The increasing penetration of multi renewables and their multi uncertainties coupling have brought great challenge to practical joint coordinated dispatch. Traditional methods such as SO and RO are not feasible due to the unavailable accurate PDF and over-conservative decisions. Based on data-driven DRO theory, this study proposes a coordinated dispatch method for the cascaded hydro-PV-pumped storage combined system. This method establishes a two-stage DRO dispatch model to formulate the daily dispatch schedule considering the complementary economic dispatch cost of the system firstly. The complementary norm constraint is introduced to limit the probability distribution confidence set and seek out the optimized adjustment scheme for the day-ahead dispatch schedule considering the adjustment cost of real-time operation under the worst distribution. The two-stage dispatch model is solved in MP-SP framework by CCG algorithm, and the optimal dispatch schedule is formed by daily dispatch schedule and adjustive dispatch scheme. The numerical dispatch results of actual demonstration area verify the effectiveness and efficiency of the proposed method.