BELTISTOS: A robust interior point method for large-scale optimal power flow problems
BELTISTOS: A robust interior point method for large-scale optimal power flow problems
复制标题
DOI：
10.1016/j.epsr.2022.108613
复制
发表时间：
2022
影响因子：
3.9
通讯作者：
J. Kardoš;D. Kourounis;O. Schenk;R. Zimmerman
中科院分区：
文献类型：
--
作者：
J. Kardoš;D. Kourounis;O. Schenk;R. Zimmerman
最优潮流（OPF）问题是电网日常运行和规划中普遍存在的问题。这些最优控制问题是非线性的，非凸的，对于大型电网来说，特别是对于定义在大量时间段内的OPF问题，计算要求很高，由于与储能设备相关的约束，这些问题通常是跨时间耦合的。提出了一个鲁棒的内点优化库BELTISTOS，该库可以快速准确地求解单周期OPF问题，并通过结构挖掘算法显著加快多周期OPF问题的求解速度。BELTISTOS坚持可复制和可靠分析的高报告标准，与软件包MATPOWER中的内部点优化器进行比较，并使用多达193,000个总线的大型电网和长达4800个时间段的问题进行评估。
Optimal power flow (OPF) problems are ubiquitous for daily power grid operations and planning. These optimal control problems are nonlinear, non-convex, and computationally demanding for large power networks especially for OPF problems defined over a large number of time periods, which are commonly intertemporally coupled due to constraints associated with energy storage devices. A robust interior point optimization library BELTISTOS is proposed, which allows fast and accurate solutions to single-period OPF problems and significantly accelerates the solution of multi-period OPF problems via the aid of structure-exploiting algorithms. Adhering to high reporting standards for replicable and reliable analysis, BELTISTOS is compared with interior point optimizers within the software package MATPOWER and evaluated using large scale power networks with up to 193,000 buses and problems spanning up to 4800 time periods.