计及多元不确定性的氢电耦合微电网多阶段动态规划方法

针对传统单阶段规划设计方法存在的初始投资成本高、源荷匹配性弱、经济性不足的缺点，提出了一种氢电耦合微电网多阶段动态规划方法，可确定各设备的最优配置容量和投资时机，并基于混合优化架构增强模型对多元不确定性的适应能力。其中，设备在规划周期内投资成本的波动不确定性特征以随机概率场景描述，可再生分布式电源出力和负荷需求不确定性则以不确定场景集合处理。在此基础上，进一步考虑微电网规划周期内的负荷增长率、设备性能衰退等动态因素，提升规划方案对未来动态信息变化的适用性，并最终构建随机-鲁棒混合整数线性规划模型描述微电网的多阶段动态规划问题。仿真结果表明，与传统的单阶段规划模型相比，所提方法的总成本降低近5.0%，初始投资成本降低近13.9 %，投资回收期缩短近32.9 %，有利于提升微电网规划方案的经济性。

Multi-stage dynamic programming method for hydrogen-electric coupled microgrid considering multiple uncertainties

Traditional single-stage planning methods face with shortcomings such as high initial investment cost, weak matching degree between sources and loads, and insufficient economy. To this end, a multi-stage dynamic planning model for hydrogen-electric coupled microgrid is proposed. Both the optimal capacity and investment time of each device can be obtained by the proposed model, and the adaptability to multiple uncertainties is enhanced based on hybrid optimization architecture. The fluctuation uncertainty feature of equipment investment cost in the planning cycle is described by stochastic probability scenarios, and the uncertainty of renewable distributed generation output and load demand is addressed by uncertainty sets. Moreover, considering the dynamic factors such as load growth rate and device performance degradation in the planning cycle of microgrid to further improve the applicability of the planning scheme to the variation of future dynamic information, and the stochastic-robust mixed integer linear programming model is constructed to describe the multi-stage dynamic programming problem in the end. Simulative results show that compared with the traditional single-stage planning model, the total cost of the proposed method is reduced by nearly 5.0%,the initial investment cost is reduced by nearly 13.9 %,and the payback period is shor-tened by nearly 32.9 %,which is conducive to improving the economy of planning scheme for microgrid.