基于主动迁移学习的电力系统暂态稳定自适应评估  

Adaptive Assessment of Power System Transient Stability Based on Active Transfer Learning

作　　者：赵晨浩 焦在滨[1] 李程昊 张迪 张鹏辉 ZHAO Chenhao;JIAO Zaibin;LI Chenghao;ZHANG Di;ZHANG Penghui(School of Electrical Engineering,Xi'an Jiaotong University,Xi'an 710049,China;Electric Power of Henan,Electric Power Research Institute,Zhengzhou 450052,China)

机构地区：[1]西安交通大学电气工程学院,陕西西安710049 [2]国网河南省电力公司电力科学研究院,河南郑州450052

出　　处：《中国电力》2025年第1期70-77,共8页Electric Power

基　　金：国家电网有限公司科技项目(5100-202124011A-0-0-00)。

摘　　要：构建了一个基于主动迁移学习的框架,基于原始场景数据搭建并训练源域暂态稳定评估(transient stability assessment,TSA)模型。当运行场景变化导致模型性能下降时启动更新机制,通过短时时域仿真生成大量无稳定性标签的样本以及完整仿真生成小批量带标签样本,采用基于变分对抗的主动学习方法学习数据潜在的特征表示空间,根据置信度选择信息量最大的无标签样本并进行标注。迁移基础模型参数并结合有标签样本进行微调,在保证迁移精度的情况下节省更新时间,通过IEEE 39节点验证了所提方法的有效性。This paper constructs a framework based on active transfer learning.The basic model is built and trained based on the original scene data.The update mechanism is started when the performance of the model decreases due to the change of the running scene.A large number of samples without stable state are generated by short-term time-domain simulation,and a small batch of labeled samples are generated by complete simulation.The active learning method based on variational adversarial is used to learn the potential feature representation space of the data,and the unlabeled samples with the largest amount of information are selected and labeled according to the confidence.The basic model parameters are migrated and fine-tuned with labeled samples to save the update time while ensuring the migration accuracy.The IEEE 39 node verifies the effectiveness of the proposed method.

关 键 词：电力系统 暂态稳定评估 迁移学习 主动学习 

分 类 号：TM712[电气工程—电力系统及自动化] TP181[自动化与计算机技术—控制理论与控制工程]