闫涛, 李蕊, 刘伟, 盛万兴, 孟晓丽. 基于改进停电序列多状态模型的新型配电系统故障恢复方案解析优选方法[J]. 电网技术, 2023, 47(2): 473-481. DOI: 10.13335/j.1000-3673.pst.2022.0185
引用本文: 闫涛, 李蕊, 刘伟, 盛万兴, 孟晓丽. 基于改进停电序列多状态模型的新型配电系统故障恢复方案解析优选方法[J]. 电网技术, 2023, 47(2): 473-481. DOI: 10.13335/j.1000-3673.pst.2022.0185
YAN Tao, LI Rui, LIU Wei, SHENG Wanxing, MENG Xiaoli. Analytic Optimization Method of Fault Recovery Scheme for New Distribution Network Based on Modified Multi-state Power Interruption Sequence Model[J]. Power System Technology, 2023, 47(2): 473-481. DOI: 10.13335/j.1000-3673.pst.2022.0185
Citation: YAN Tao, LI Rui, LIU Wei, SHENG Wanxing, MENG Xiaoli. Analytic Optimization Method of Fault Recovery Scheme for New Distribution Network Based on Modified Multi-state Power Interruption Sequence Model[J]. Power System Technology, 2023, 47(2): 473-481. DOI: 10.13335/j.1000-3673.pst.2022.0185

基于改进停电序列多状态模型的新型配电系统故障恢复方案解析优选方法

Analytic Optimization Method of Fault Recovery Scheme for New Distribution Network Based on Modified Multi-state Power Interruption Sequence Model

  • 摘要: 新能源不确定性引发的场景维数灾是新型配电系统优化分析的核心问题,但穷举搜索运算量大且耗时过长,随机优化则存在局部最优与收敛困难两大难题。鉴于此,提出了一种基于改进停电序列多状态模型的新型配电系统故障恢复方案解析优选方法,可通过降低时间复杂度来解决场景维数灾。首先,建立了源荷储多状态模型及其耦合运行状态矩阵;然后,改进停电序列多状态模型并给出了状态概率计算方法;最后,提出了新型配电系统的故障恢复解析优选算法。算例分析中,所提方法不仅能在15min内完成千垓级(1023)场景的方案解析评估,而且比穷举搜索和随机优化具有更明显的算法性能优势。此外,算例还分析了源荷储状态的离散程度对算法精度和计算速度的影响。

     

    Abstract: This paper proposes a fault recovery scheme analytic optimization method for a new distribution network based on the modified multi-state power interruption sequence model, which may solve the scenario dimension disaster by reducing the time complexity. Firstly, the multi-state model of source-load-storage and their coupled operation state matrix are established. Then, the multi-state power interruption sequence model is modified and the calculation method of its state probability is given. Finally, the fault recovery scheme analytical optimization algorithm for the new distribution system is proposed. The illustrative example shows that the proposed method evaluates analytically a recovery scheme with 1023 different scenarios within 15min, and has an obvious algorithm performance advantage compared with the exhaustive search method and the stochastic optimization method. In addition, the effect of the dispersion degree of the generation-load-storage state on the precision and speed of the proposed method is analyzed.

     

/

返回文章
返回