浙江电力

2026, v.45;No.366(10) 79-90

[打印本页] [关闭]
本期目录(Current Issue) | 过刊浏览(Archive) | 高级检索(Advanced Search)

基于改进Stacking集成学习的台区电压越限评估
Voltage violation evaluation for distribution transformer areas based on modified Stacking ensemble learning

孔腊,杨秀,孙改平,杨帆
KONG La,YANG Xiu,SUN Gaiping,YANG Fan

摘要(Abstract):

针对分布式光伏在配电网中大规模接入后,配电台区的节点电压越限问题,提出了基于改进Stacking集成学习的台区电压越限评估方法。首先,基于实际拓扑分析得到不同光伏渗透率、光伏接入位置、功率因数、负载率场景下对应的节点电压,构建电压越限严重度指标并计算不同场景下对应的电压越限严重度。其次,基于XGboost(极端梯度提升)评估方法,量化各因素影响电压越限的程度,得出低压台区电压越限的主要成因,并将各成因作为输入数据。最后,基于改进Stacking集成学习的方法,评估低压台区越限的严重度。测试结果表明,所提方法准确率高,为治理台区提供了支撑,可提高供电安全性。
To address the node voltage violation within distribution transformer areas following the large-scale integration of distributed photovoltaic(PV) systems into distribution networks, this paper presents a voltage violation severity evaluation method based on feature-weighted Stacking ensemble learning. First, node voltages are obtained under various scenarios with distinct PV penetration levels, PV connection locations, power factors, and load rates based on actual topology analysis. A voltage violation severity index is formulated to calculate the corresponding severity values for these different scenarios. Subsequently, an XGBoost-based evaluation method is deployed to quantify the impact of each factor on voltage violations, identifying the primary causes of voltage violations in low-voltage distribution transformer areas to serve as input variables. Finally, the severity of voltage violations in low-voltage distribution transformer areas is evaluated using the modified Stacking ensemble learning. Test results demonstrate that the proposed method achieves high evaluation accuracy, providing support for distribution transformer area regulation and enhancing the security of power supply.

关键词(KeyWords): 低压台区电压;电压越限;严重度评估;XGboost;Stacking集成学习
voltage of low-voltage distribution transformer area;voltage violation;severity evaluation;XGBoost;Stacking ensemble learning

Abstract:

Keywords:

基金项目(Foundation): 国家自然科学基金(52177098);; 上海电力人工智能工程技术研究中心项目(19DZ2252800)

作者(Author): 孔腊,杨秀,孙改平,杨帆
KONG La,YANG Xiu,SUN Gaiping,YANG Fan

DOI: 10.19585/j.zjdl.202610008

参考文献(References):

扩展功能
本文信息
服务与反馈
本文关键词相关文章
本文作者相关文章
中国知网
分享