摘要
随着碳中和目标的提出以及分布式可再生能源的发展,微电网作为未来集成清洁能源的有效载体受到广泛关注。风电等可再生能源的随机性和间歇性给微电网的经济调度带来挑战,为了减少不确定性的影响,基于数据驱动的条件正态Copula函数对风电预测误差的时间相关性及其与日前预测值的条件相关性进行建模,通过K-means聚类得到给定的日前预测值下第二日可能的风电功率场景,并采用随机优化方法解决微电网的日前调度问题。仿真结果表明,所提方法能够合理反映日前风电功率的可能场景,有效降低微电网运行成本,相较于确定性优化模型,日内不平衡电量减少22.77%,运行成本降低4.01%。此外,相较于不考虑相关性的传统随机优化模型,日内不平衡电量减少18.23%,运行成本降低3.2%。
Considering the goal of carbon neutralization,and the development of distributed renewable energy,microgrids have attracted wide attention as effective carriers of integrated clean energy in the future.The randomness,and intermittence of renewable energy,such as wind power,pose challenges for the economic dispatching of microgrids.To reduce the influence of uncertainty,in this study,we modeled a time-correlation of wind power prediction error,and its conditional correlation with day-ahead forecast,based on data-driven conditional normal copula.We obtained possible wind power scenarios under day-ahead prediction by using K-means clustering,and uses used a stochastic optimization method to solve the dayahead scheduling problem of microgrids with energy storage.The simulation results showed that the proposed method could reasonably reflect the possible scenarios of day-ahead wind power,and effectively reduce the operating cost of the microgrid.Compared with the deterministic optimization model,the intraday unbalanced power was reduced by 22.77%,and the operating cost was reduced by 4.01%.Additionally,compared with the traditional stochastic optimization model that does not consider correlation,the intraday unbalanced power was reduced by 18.23%,and the operating cost was reduced by 3.2%.
作者
姜宇
陈翔宇
傅守强
JIANG Yu;CHEN Xiangyu;FU Shouqiang(State Grid JIBEI Electric Power Company Limited,Xicheng District,Beijing 100054,China)
出处
《全球能源互联网》
CSCD
2022年第1期46-54,共9页
Journal of Global Energy Interconnection
基金
国网冀北电力有限公司科技项目(52018F20001M)。
