计及风电与电动汽车随机性的两阶段机组组合研究

风电出力的随机性以及电动汽车(electric vehicle,EV)充电需求的不确定性给电力系统调度带来了挑战。在传统确定性机组组合模型的基础上,针对电力系统日前调度面临的不确定问题,提出了充分考虑风电与电动汽车双重不确定性的随机优化调度及备用计算模型。首先,对于风电出力不确定性,采用基于场景分析的两阶段随机优化方法,并使用生成对抗网络(generative adversarial network, GAN)来生成风电场景。其次,对于电动汽车充电需求的不确定性,将其分为可调度与不可调度两类。可调度电动汽车根据其出行规律采用随机模拟的方法,并建立了EV充电聚集商模型;不可调度电动汽车通过K-means聚类分析得到其典型负荷曲线,并将其并入系统常规负荷中。最终建立了基于多场景分析考虑EV充电聚集商的两阶段随机机组组合模型,并通过算例分析证明了所提模型的有效性。

Two-Stage Stochastic Unit Commitment Considering the Uncertainty of Wind Power and Electric Vehicle Travel Patterns

The uncertainty of wind power output and charging demand of electric vehicles (EVs) brings great challenge for power system dispatching. Taking day-ahead dispatch as the target, this paper proposes a stochastic optimization scheduling model and corresponding reserve model on the basis of traditional unit commitment model. Firstly, for the uncertainty of wind power output, a two-stage stochastic unit commitment model is established in this paper according to scenario analysis based on generative adversarial network (GAN), while electric vehicles are divided into two categories: schedulable and non-schedulable EVs. Monte-Carlo method is adopted to simulate the behavior and the dynamic change of schedulable EVs on the basis of probability distribution of the travel patterns, and the model of EV aggregators is established in this paper. As for non-schedulable EVs, K-means cluster analysis is adopted to get a typical load curve, and then the charging demand is viewed as part of conventional load. Case study demonstrates the validity of the proposed model.