基于主动迁移学习的电力系统暂态稳定自适应评估

doi:10.11930/j.issn.1004-9649.202404033

摘要/Abstract

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。

关键词: 电力系统, 暂态稳定评估, 迁移学习, 主动学习

Abstract:

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.

Key words: power system, transient stability assessment, transfer learning, active learning

中图分类号:

TM732

赵晨浩, 焦在滨, 李程昊, 张迪, 张鹏辉. 基于主动迁移学习的电力系统暂态稳定自适应评估[J]. 中国电力, 2025, 58(1): 70-77.

Chenhao ZHAO, Zaibin JIAO, Chenghao LI, Di ZHANG, Penghui ZHANG. Adaptive Assessment of Power System Transient Stability Based on Active Transfer Learning[J]. Electric Power, 2025, 58(1): 70-77.

导出引用管理器 EndNote|Ris|BibTeX

链接本文: https://www.electricpower.com.cn/CN/10.11930/j.issn.1004-9649.202404033

https://www.electricpower.com.cn/CN/Y2025/V58/I1/70

图/表 12

图 1 TCN-GRU模型结构

Fig.1 The structure of TCN-GRU model

图 2 主动学习流程

Fig.2 The process of active learning

图 3 变分对抗主动学习

Fig.3 The architecture of variational adversarial active learning

图 4 基于主动迁移学习的TSA框架

Fig.4 The TSA framework based on active transfer learning

图 5 IEEE-39节点电力系统

Fig.5 The IEEE 39-bus power system

表 1 混淆矩阵

Table 1 Confusion matrix

真实标签	预测标签
真实标签	稳定	失稳
稳定	${T_{\text{P}}}$	${F_{\text{N}}}$
失稳	${F_{\text{P}}}$	${T_{\text{N}}}$

表 2 不同模型的评估结果对比

Table 2 Comparison of evaluation results of different models 单位：%

模型	准确率	召回率	精确率	F1分数
LR	93.23	94.52	94.67	94.59
SVM	95.82	97.02	96.31	96.67
XGBoost	96.99	98.01	97.20	97.61
DNN	96.96	96.81	98.38	97.59
CNN	96.41	97.57	96.65	97.11
TCN	97.12	97.85	97.55	97.70
GRU	96.67	97.09	97.59	97.34
TCN-GRU	98.32	99.28	98.67	98.98

表 3 源域模型在新场景下的测试结果

Table 3 The test results of the source domain model in the new scenario 单位：%

目标系统	准确率	召回率	精确率	F1分数
1	82.51	85.95	93.21	89.43
2	76.11	81.65	89.22	85.27
3	80.30	85.07	91.27	88.06

表 4 VAE结构

Table 4 The architecture of VAE

模型	层	输入通道	输出通道	卷积核大小/步长
编码器	Conv1	190	128	3/1
	Conv1	128	256	3/1
	Conv1	256	512	3/1
解码器	Deconv1	512	256	3/1
	Deconv1	256	128	3/1
	Deconv1	128	190	3/1

图 6 主动迁移学习迭代更新曲线

Fig.6 Active transfer learning iterative update curve

图 7 目标域主动迁移学习准确率变化

Fig.7 The accuracy change of active transfer learning in target domain

表 5 各方案下的测试结果

Table 5 The test results under each scheme

方案	仿真时间/s	训练时间/s	总时间/s	准确率/%
1	2400	124	2524	97.95
2	2400	16	2416	95.96
3	4000	27	4027	97.98
4	4000	346	4346	97.51

参考文献 24

1	娄奇鹤, 李荣盛, 谭捷, 等. 基于卷积神经网络的暂稳极限功率计算[J]. 中国电力, 2024, 57 (4): 211- 219.
	LOU Qihe, LI Rongsheng, TAN Jie, et al. Calculation of transient stability limit based on convolutional neural network[J]. Electric Power, 2024, 57 (4): 211- 219.
2	周生存, 罗毅, 易煊承, 等. 考虑数据缺失的图注意力网络暂态稳定评估[J]. 中国电力, 2024, 57 (5): 157- 167.
	ZHOU Shengcun, LUO Yi, YI Xuancheng, et al. Transient stability assessment of graph attention networks considering data missing[J]. Electric Power, 2024, 57 (5): 157- 167.
3	向川, 陈鎏凯, 陈勇, 等. 基于深层级联残差图卷积的暂态稳定评估模型及其实际电网应用[J]. 广东电力, 2024, 37 (6): 62- 69.
	XIANG Chuan, CHEN Liukai, CHEN Yong, et al. Application of transient stability assessment model based on deep cascading residual graph convolution network in real world power grids[J]. Guangdong Electric Power, 2024, 37 (6): 62- 69.
4	陈水耀, 胡晨, 马伟, 等. 基于故障信息的风机并网系统暂态稳定分析方法[J]. 浙江电力, 2023, 42 (9): 89- 98.
	CHEN Shuiyao, HU Chen, MA Wei, et al. A Transient stability analysis method for power systems with wind turbines integrated based on fault information[J]. Zhejiang Electric Power, 2023, 42 (9): 89- 98.
5	于枋彤, 刘铖, 张宇驰. 输电线开断控制对交直流混联电力系统暂态稳定性影响研究[J]. 东北电力大学学报, 2023, 43 (1): 85- 91.
	YU Fangtong, LIU Cheng, ZHANG Yuchi. Research on influence of transmission line breaking control on transient stability of AC/DC hybrid power system[J]. Journal of Northeast Electric Power University, 2023, 43 (1): 85- 91.
6	李宇骏, 陆艺源, 牟同鹏, 等. 新能源发电经换流器并网系统的稳定性分析与控制综述[J]. 电网与清洁能源, 2023, 39 (12): 79- 94.
	LI Yujun, LU Yiyuan, MU Tongpeng, et al. A review of stability analysis and control of renewable energy integration systems via converters[J]. Power System and Clean Energy, 2023, 39 (12): 79- 94.
7	曾垂辉, 谢俊, 李智, 等. 探究无功控制回路对构网型并网逆变器暂态稳定性的影响[J]. 东北电力大学学报, 2024, 44 (4): 38- 45.
	ZENG Chuihui, XIE Jun, LI Zhi, et al. Investigating the impact of reactive power control loops on the transient stability of grid-forming inverters[J]. Journal of Northeast Electric Power University, 2024, 44 (4): 38- 45.
8	伏绍鑫, 张路, 唐翰峰, 等. 考虑柔性电热负荷的区域综合能源系统低碳经济调度[J]. 电力科技与环保, 2023, 39 (5): 417- 428.
	FU Shaoxin, ZHANG Lu, TANG Hanfeng, et al. Low-carbon economic dispatch of community integrated energy system considering flexible electric heating load[J]. Electric Power Technology and Environmental Protection, 2023, 39 (5): 417- 428.
9	刘昭睿, 王凯, 王新建, 等. 提升末端区域电网供电能力的调控方案及稳控策略研究[J]. 内蒙古电力技术, 2024, 42 (3): 52- 60.
	LIU Zhaorui, WANG Kai, WANG Xinjian, et al. Research on regulation scheme and stability control strategy for enhancing power supply capacity of end-region power grid[J]. Inner Mongolia Electric Power, 2024, 42 (3): 52- 60.
10	王中行, 周元贵, 张学广. 基于人工智能算法的风电机组状态监测和故障诊断技术研究综述[J]. 东北电力大学学报, 2024, 44 (1): 42- 51.
	WANG Zhongxing, ZHOU Yuangui, ZHANG Xueguang. Review of artificial intelligence algorithms-based wind turbine condition monitoring and fault diagnosis techniques[J]. Journal of Northeast Electric Power University, 2024, 44 (1): 42- 51.
11	李子凯, 岳宝强, 杨波, 等. 适于低功率状态的多特征融合负荷分解方法[J]. 山东电力技术, 2024, 51 (1): 68- 76.
	LI Zikai, YUE Baoqiang, YANG Bo, et al. Multi-feature fusion load disaggregation method for low-power states[J]. Shandong Electric Power, 2024, 51 (1): 68- 76.
12	姜力杨, 盖晨昊, 齐航, 等. 数据驱动的电力系统动态安全评估研究综述[J]. 山东电力技术, 2024, 51 (4): 27- 35.
	JIANG Liyang, GAI Chenhao, QI Hang, et al. Review on data-driven dynamic security assessment of power systems[J]. Shandong Electric Power, 2024, 51 (4): 27- 35.
13	邢法财, 麻常辉, 苗伟威, 等. 考虑自然灾害影响的节点恢复重要度评估与目标网架重构[J]. 浙江电力, 2024, 43 (10): 53- 64.
	XING Facai, MA Changhui, MIAO Weiwei, et al. Assessment of the importance of node recovery and grid framework reconfiguration with consideration of natural disaster impacts[J]. Zhejiang Electric Power, 2024, 43 (10): 53- 64.
14	武永强, 郭凯, 王宇强, 等. 构网型储能系统并网测试要求及测试实例分析[J]. 内蒙古电力技术, 2024, 42 (2): 8- 16.
	WU Yongqiang, GUO Kai, WANG Yuqiang, et al. Testing requirements and case analysis of grid-forming energy storage system[J]. Inner Mongolia Electric Power, 2024, 42 (2): 8- 16.
15	王汝田, 王世琪. 分裂源双级矩阵变换器及其调制策略的研究[J]. 东北电力大学学报, 2023, 43 (3): 47- 54.
	WANG Rutian, WANG Shiqi. Research on split-source two-stage matrix converter and its modulation strategy[J]. Journal of Northeast Electric Power University, 2023, 43 (3): 47- 54.
16	李博, 石红晖, 马强, 等. 风电-火电-压缩空气储能综合能源系统运行特性研究[J]. 电力科技与环保, 2024, 40 (2): 168- 177.
	LI Bo, SHI Honghui, MA Qiang, et al. The operating characteristics research of integrated energy system based on the wind, coal-fired power plant and compressed air energy storage[J]. Electric Power Technology and Environmental Protection, 2024, 40 (2): 168- 177.
17	孙黎霞, 白景涛, 周照宇, 等. 基于双向长短期记忆网络的电力系统暂态稳定评估[J]. 电力系统自动化, 2020, 44 (13): 64- 72.
	SUN Lixia, BAI Jingtao, ZHOU Zhaoyu, et al. Transient stability assessment of power system based on bi-directional long-short-term memory network[J]. Automation of Electric Power Systems, 2020, 44 (13): 64- 72.
18	贺日星, 杨宝峰, 陈瀚栋, 等. 抑制直流输电换相失败方法研究[J]. 内蒙古电力技术, 2023, 41 (6): 75- 83.
	HE Rixing, YANG Baofeng, CHEN Handong, et al. Research of methods for suppressing commutation failure in HVDC transmission[J]. Inner Mongolia Electric Power, 2023, 41 (6): 75- 83.
19	陈哲, 李山林, 林达, 等. 考虑拓扑切换暂态波动的微电网群分布式协同控制策略[J]. 浙江电力, 2023, 42 (11): 1- 10.
	CHEN Zhe, LI Shanlin, LIN Da, et al. A distributed cooperative control strategy for MGC considering transient fluctuations in topology switching[J]. Zhejiang Electric Power, 2023, 42 (11): 1- 10.
20	张若愚, 吴俊勇, 李宝琴, 等. 基于迁移学习的电力系统暂态稳定自适应预测[J]. 电网技术, 2020, 44 (6): 2196- 2205.
	ZHANG Ruoyu, WU Junyong, LI Baoqin, et al. Self-adaptive power system transient stability prediction based on transfer learning[J]. Power System Technology, 2020, 44 (6): 2196- 2205.
21	XIE J, SUN W. A transfer and deep learning-based method for online frequency stability assessment and control[J]. IEEE Access, 2021, 9, 75712- 75721. DOI
22	汤奕, 崔晗, 党杰. 基于继承思想的时变性电力系统暂态稳定预测[J]. 中国电机工程学报, 2021, 41 (15): 5107- 5119.
	TANG Yi, CUI Han, DANG Jie. Transient stability prediction of time-varying power systems based on inheritance[J]. Proceedings of the CSEE, 2021, 41 (15): 5107- 5119.
23	焦昊, 殷岩岩, 吴晨, 等. 基于安全强化学习的主动配电网有功-无功协调优化调度[J]. 中国电力, 2024, 57 (3): 43- 50.
	JIAO Hao, YIN Yanyan, WU Chen, et al. Coordinated optimization of active and reactive power of active distribution network based on safety reinforcement learning[J]. Electric Power, 2024, 57 (3): 43- 50.
24	YAN R, WANG Z Y, YUAN Y X, et al. Information entropy based prioritization strategy for data-driven transient stability batch assessment[J]. CSEE Journal of Power and Energy Systems, 2021, 7 (3): 443- 455.

[1]	李健, 张钧, 韩新阳, 靳晓凌. 新型电力系统形态量化推演方法的总体框架与功能设计[J]. 中国电力, 2025, 58(3): 1-7, 97.
[2]	张磊, 马晓伟, 王满亮, 陈力, 高丙团. 互联新能源电力系统区内AGC机组分布式协同控制策略[J]. 中国电力, 2025, 58(3): 8-19.
[3]	张沛, 杨马婧, 张放, 谢桦, 刘广一, 李文云, 路学刚, 王珍意, 翟苏巍. 图数据库与图计算在电力系统调度运行应用综述[J]. 中国电力, 2025, 58(3): 119-131.
[4]	高志远, 庄卫金, 李峰, 于芳, 张鸿, 王艳, 夏旻. 调控领域人工智能应用的高复用性验证平台[J]. 中国电力, 2025, 58(3): 142-150.
[5]	汤明润, 李若旸, 刘慕然, 程晓钰, 刘铫, 杨淑霞. 电力系统稳态下可再生能源大规模接入量预测[J]. 中国电力, 2025, 58(2): 126-132.
[6]	任明, 李乾宇, 夏昌杰, 董明. 基于稀疏光谱成像的线路复合绝缘子积污状态可视化评估[J]. 中国电力, 2025, 58(2): 203-215.
[7]	董武, 张健, 周勤勇, 张立波, 龚浩岳. 中国电力系统安全稳定性演化综述[J]. 中国电力, 2025, 58(1): 115-127.
[8]	朱瑞, 娄奇鹤, 靳晓凌, 刘畅. 适应新型电力系统的电网基建功能化投资结构演化研究[J]. 中国电力, 2024, 57(9): 194-204.
[9]	傅观君, 张富强, 夏鹏, 冯君淑, 张晋芳. 天然气发电在新型电力系统中的功能定位及发展前景研判[J]. 中国电力, 2024, 57(8): 67-74.
[10]	翟苏巍, 李银银, 杜凡, 李文云, 潘凯岩, 梁峻恺. 考虑海量分布式能源接入的配电网分布式无功控制策略[J]. 中国电力, 2024, 57(8): 138-144.
[11]	李碧君, 李威, 董希建. 新型电力系统中主控型安全稳定服务质量的在线评估[J]. 中国电力, 2024, 57(8): 168-181.
[12]	周生存, 罗毅, 易煊承, 吴亚宁, 李丁, 熊逸. 考虑数据缺失的图注意力网络暂态稳定评估[J]. 中国电力, 2024, 57(5): 157-167.
[13]	叶小宁, 王彩霞, 李琼慧, 杨超. 国外新能源高占比电力系统电力供应保障措施及启示[J]. 中国电力, 2024, 57(4): 61-67.
[14]	王雪, 刘林, 卓庆东, 张海鹏, 杨苓, 许方园, 曹雨晨. 基于虚拟惯量控制的新型电力系统功率差前馈振荡抑制方法[J]. 中国电力, 2024, 57(4): 68-76.
[15]	周勤勇, 李根兆, 秦晓辉, 施浩波, 陈文静, 龚浩岳. 能源革命下的电力系统范式转换分析[J]. 中国电力, 2024, 57(3): 1-11.