A stochastic dual dynamic programming approach for optimal operation of DER aggregators
A stochastic dual dynamic programming approach for optimal operation of DER aggregators
复制标题
DOI：
10.1109/ptc.2017.7981213
复制
发表时间：
2017-06
期刊：
影响因子：
--
通讯作者：
Panagiotis Fatouros;I. Konstantelos;D. Papadaskalopoulos;G. Strbac
中科院分区：
文献类型：
--
作者：
Panagiotis Fatouros;I. Konstantelos;D. Papadaskalopoulos;G. Strbac
分布式能源（DER）聚合器的运行是一项高度复杂的任务，受许多不确定因素的影响，如可再生能源注入，负荷水平和市场条件。然而，传统的随机规划方法忽略了周围的不确定变量的时间依赖性的信息，由于计算易处理的限制。本文提出了一种新的随机对偶动态规划（SDDP）的DER聚合器的最优操作方法。传统的SDDP框架进行了扩展，捕捉时间依赖的不确定性的风力发电输出，通过集成的n阶自回归（AR）模型。这种方法被证明是实现一个更好的解决方案的效率和计算时间的要求相比，传统的随机规划方法的基础上使用的情况树之间的权衡。
The operation of aggregators of distributed energy resources (DER) is a highly complex task that is affected by numerous factors of uncertainty such as renewables injections, load levels and market conditions. However, traditional stochastic programming approaches neglect information around temporal dependency of the uncertain variables due to computational tractability limitations. This paper proposes a novel stochastic dual dynamic programming (SDDP) approach for the optimal operation of a DER aggregator. The traditional SDDP framework is extended to capture temporal dependency of the uncertain wind power output, through the integration of an n-order autoregressive (AR) model. This method is demonstrated to achieve a better trade-off between solution efficiency and computational time requirements compared to traditional stochastic programming approaches based on the use of scenario trees.