基于两阶段迁移学习的电力系统暂态稳定评估框架  

Transient Stability Assessment Framework of Power System Based on Two-stage Transfer Learning

作　　者：李保罗 孙华东 张恒旭[1] 高磊 徐式蕴 黄彦浩 LI Baoluo;SUN Huadong;ZHANG Hengxu;GAO Lei;XU Shiyun;HUANG Yanhao(Key Laboratory of Power System Intelligent Dispatch and Control of Ministry of Education(Shandong University),Jinan 250061,China;China Electric Power Research Institute Co.,Ltd.,Beijing 100192,China)

机构地区：[1]电网智能化调度与控制教育部重点实验室(山东大学),山东省济南市250061 [2]中国电力科学研究院有限公司,北京市100192

出　　处：《电力系统自动化》2022年第17期176-185,共10页Automation of Electric Power Systems

基　　金：国家重点研发计划资助项目(响应驱动的大电网稳定性智能增强分析与控制技术,2021YFB2400800)。

摘　　要：为提高基于数据驱动的暂态稳定评估模型对电网的自适应性,将深度迁移学习引入更新过程,提出一种基于两阶段迁移学习的暂态稳定评估框架。所提框架根据时间尺度分为2个阶段:在第1阶段,利用深度子领域自适应网络挖掘无标注数据信息,将模型的评估性能快速提升到相对可靠的水平,得到迁移模型,结合时域仿真法进行综合判稳,提高电网变化初期模型的可用性;在第2阶段,利用迁移模型筛选高价值样本集,并结合样本迁移和微调技术进行二次更新,使评估性能恢复到较高水平,降低更新时间成本。在IEEE 39节点系统和中国某省级电网模型上进行测试,结果表明所提框架具有完备性,可使模型快速响应电网的变化。In order to improve the adaptability of the data-driven transient stability assessment model to the power grid,deep transfer learning is introduced into the updating process,and a transient stability assessment framework based on two-stage transfer learning is proposed.The proposed framework is divided into two stages according to the time scale.In the first stage,the deep subdomain adaptive network is used to mine unlabeled data information,and the evaluation performance of the model is rapidly improved to a relatively reliable level.The obtained transfer model is combined with the time domain simulation method for comprehensive stability judgment,which improves the usability of the model in the initial stage of power grid change.In the second stage,the transfer model is used to select the high-value sample set,and the second update is carried out by combining the sample transfer and fine-tuning technology to restore the evaluation performance to a higher level and reduce the updating time cost.The test results in the models of IEEE 39-bus system and a provincial power grid in China show that the proposed framework has completeness and can make the model quickly respond to the changes in the power grid.

关 键 词：暂态稳定 评估 深度迁移学习 数据驱动应 

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]