考虑风电不确定性的电-气能源系统数据驱动鲁棒优化调度

A Data-driven Robust Unit Commitment Model of Integrated Electricity and Gas System Considering Wind Power Uncertainty

电-气能源系统的耦合为消纳风电提供了新途径,但风电随机过程的复杂不确定性给电-气能源系统运行带来了挑战。基于此,提出了一种考虑风电预测误差不确定性的电-气能源系统数据驱动鲁棒优化调度模型。首先通过无穷维高斯混合模型对风电预测误差进行聚类,考虑风电厂间出力的相关性,建立了数据驱动的风电预测误差不确定集。接着以燃气轮机为耦合元件,考虑动态天然气潮流,提出了两阶段的电-气能源系统鲁棒优化调度模型。第一阶段为日前计划层面,决策机组启停及出力区间、天然气的供气量区间;第二阶段为实时运行层面,决策机组的出力值以及天然气的供气量,针对天然气网络的非凸非线性约束,采用二阶锥松弛的方法将其进行凸化。针对该模型min-max-min的三层结构,通过KKT条件将第二阶段问题转化为单层问题,然后采用列约束生成算法进行求解。最后,通过6节点与118节点的电-气能源系统的算例分析,证明了所提模型的有效性。

The integrated electricity and gas system(IEGS) provides a new way to increase wind power consumption. In order to cope with the complex uncertainty of wind power generation to the system operation, a data-driven robust unit commitment of IEGS considering the uncertainty of wind power generation is proposed in this paper. Firstly, wind power forecast errors are clustered by using the infinite Gaussian mixture model, and a data-driven uncertainty set is derived afterwards. Then a two-staged IEGS unit commitment model is established in which the gas-fired units are coupled into one and the dynamic natural gas flow equations are considered. In the first stage, the unit status, the power bounds, and the gas supplement bounds are decided, while in the second stage, the power output and the amount of natural gas supply are decided. The non-convex constraints of the dynamic natural gas flow are transformed into convex ones by the second-order cone technique. In order to solve this min-max-max-min problem, a Column Constraint Generation(C&CG) based algorithm is introduced. Finally, the effectiveness of the proposed model is verified by a 6-6 bus system and a 108-10 bus system.