• 主题
高级搜索  >
历史搜索 清空历史
首页> 外文期刊>Electric Power Systems Research >A hybrid data-driven robust optimization approach for unit commitment considering volatile wind power
【24h】

A hybrid data-driven robust optimization approach for unit commitment considering volatile wind power

机译：A hybrid data-driven robust optimization approach for unit commitment considering volatile wind power

获取原文
获取原文并翻译 | 示例
           
页面导航
  • 摘要
  • 著录项
  • 相关主题

摘要

This paper presents a data-driven hybrid method combining kernel density estimation (KDE) and Wasserstein metric to solve unit commitment problems considering volatile wind power. The proposed approach adopts KDE to deduce the underlying probability distribution function (PDF) of wind power forecasting errors, and utilizes the confidence interval of the estimated PDF to construct the support of uncertainties. Wasserstein metric is applied to measure the distances among distributions belonging to the established support. We take the PDF estimated by KDE as the center, and take a certain Wasserstein distance, so as to form a Wasserstein ball of probability distributions, namely ambiguity set. The unit commitment plan is obtained by minimizing the expected cost with regard to the worst-case distribution in the ambiguity set. As a fusion of Wasserstein technique and KDE, the proposed method inherits good abilities of conventional Wasserstein-based method, such as statistics, tractability, extensibility and data mining capacity. More importantly, since the PDF estimated by KDE can converge to the true distribution with the increase of sample data, the proposed method can effectively further narrow the conservatism of the model without losing robustness. Computational results on IEEE 57-bus, 118-bus and 145-bus systems from MATPOWER 6.0 numerically demonstrate the favorable features of the proposed method.

著录项

获取原文

客服邮箱：kefu@zhangqiaokeyan.com

京公网安备：11010802029741号 ICP备案号：京ICP备15016152号-6 六维联合信息科技 (北京) 有限公司©版权所有

  • 客服微信

  • 服务号