考虑场景聚类的配电网−天然气联合系统双层随机运行优化
摘要：风电大规模接入配电网给系统安全经济运行带来了极大挑战。为充分表征风电随机性和不确定性对系统运行优化的影响，提出了考虑场景聚类的配电网−天然气联合系统双层随机运行优化模型。首先，通过主成分分析法对海量的高维度风电出力场景进行降维，在降维数据的基础上构建了基于分层聚类算法的风电随机场景选取方法，并提出最优聚类类数的确定标准以有效划分风电场景类数。其次，构建了基于分层场景聚类法的配电网−天然气网耦合系统双层随机运行优化模型，从多个时间尺度提高运行方案对风电波动的适应性。同时，引入拉格朗日因子，并提出KKT（Karush-Kuhn-Tucker）理论的等价方程表示方法，将所提随机运行优化模型转化为单层求解。通过算例，将所提方法与现有运行优化策略对比，验证了所提方法的有效性和优越性。
Abstract:Large-scale connection of wind power into distribution network brings great challenge to the secure and economic operation of the system. To fully characterize the impact of randomness and uncertainty of wind power on distribution system, firstly, by means of principal component analysis the dimensions of massive high-dimensional scenario of large amounts of wind power output were reduced, on this basis a stochastic scenario selection method of wind power, which was based on hierarchical clustering algorithm, was constructed, and the standard to determine the number of optimal clusterings was proposed to effectively divide the categories of wind power scenarios. Secondly, a hierarchical scenarios clustering method-based two-layer stochastic operation optimization system model for distribution network-natural gas network coupling system was built to improve the adaptability of operation scheme to wind power fluctuation from multiple time scales. Meanwhile, the Lagrange factor was led in, and the equivalent equation representation method of Karush–Kuhn–Tucker theory was presented to translate the built stochastic operation optimization model into single layer optimization problem to solve. By means of computing example, the proposed method is compared with existing operation optimization strategy, thereby the effectiveness and superiority of the proposed method are verified.
标题：考虑场景聚类的配电网−天然气联合系统双层随机运行优化
英文标题：Bi-level Stochastic Operation Optimization of Distribution- Natural Gas Combined System Considering Scenario Clustering
作者：唐海国, 张志丹, 康童, 张帝, 张聪, 罗波,
关键词：天然气网, 配电网, 场景聚类, 运行优化, 随机调度,
发表日期：2021-12-10
- 文件大小:
- 2.83 MB
- 下载次数:
- 60
-
高速下载
|
|
