部分可观测条件下的配电网数据驱动最优潮流模型

doi:10.11930/j.issn.1004-9649.202306041

摘要/Abstract

摘要：

电力系统最优潮流计算是典型的非线性非凸问题，线性化潮流模型主要用于将原始最优潮流问题转化为凸优化问题。配电网覆盖范围广，设备众多，模型参数维护困难，已有的基于数据驱动的线性化潮流模型多基于完备的系统量测数据，而实际中考虑经济性安装的测量单元，无法覆盖所有设备，系统量测通常是部分可观测的。为解决量测的部分可观测性问题，提出一种数据驱动线性潮流模型，并基于此构建基于数据驱动的线性化最优潮流模型，该模型对量测中的不良数据具有鲁棒性。通过对不同部分可观测场景的测试，验证了所提模型的有效性。

关键词: 数据驱动, 线性化最优潮流, 部分可观测

Abstract:

The linearized power flow (PF) model is mainly used to make the optimal power flow (OPF) problem convex. However, existing data-driven linear PF models are mostly based on complete system measurement data. Moreover, the systems are usually partially observable due to limited measuring devices for economical installation. This paper addresses the partial observability issue by proposing a data-driven linear PF model, which can be embedded in OPF. The model is robust against bad data in measurements, with its accuracy verified by numerical tests.

Key words: data-driven, linear optimal power flow (OPF) model, partial observability

李鹏华, 宋卓然, 吴文传. 部分可观测条件下的配电网数据驱动最优潮流模型[J]. 中国电力, 2023, 56(12): 51-57.

Penghua LI, Zhuoran SONG, Wenchuan WU. A Data-Driven Optimal Power Flow Model under Partial Observability[J]. Electric Power, 2023, 56(12): 51-57.

导出引用管理器 EndNote|Ris|BibTeX

链接本文: https://www.electricpower.com.cn/CN/10.11930/j.issn.1004-9649.202306041

https://www.electricpower.com.cn/CN/Y2023/V56/I12/51

图/表 13

参考文献 26

1	曾琮, 黄强, 陈德, 等. 基于改进纵横交叉算法的电网最优潮流计算[J]. 中国电力, 2021, 54 (9): 9- 16.
	ZENG Cong, HUANG Qiang, CHEN De, et al. Optimal power flow calculation with improved crisscross optimization algorithm[J]. Electric Power, 2021, 54 (9): 9- 16.
2	董昱, 董存, 于若英, 等. 基于线性最优潮流的电力系统新能源承载能力分析[J]. 中国电力, 2022, 55 (3): 1- 8.
	DONG Yu, DONG Cun, YU Ruoying, et al. Renewable energy capacity assessment in power system based on linearized OPF[J]. Electric Power, 2022, 55 (3): 1- 8.
3	王娟娟, 王涛, 刘子菡, 等. 考虑风电和负荷不确定性的输电网多目标柔性规划[J]. 中国电力, 2022, 55 (1): 168- 177.
	WANG Juanjuan, WANG Tao, LIU Zihan, et al. Multi-objective flexible planning of transmission network considering wind power and load uncertainties[J]. Electric Power, 2022, 55 (1): 168- 177.
4	LEI X Y, YANG Z F, YU J, et al. Data-driven optimal power flow: a physics-informed machine learning approach[J]. IEEE Transactions on Power Systems, 2021, 36 (1): 346- 354. DOI
5	TAN Y, CHEN Y Y, LI Y, et al. Linearizing power flow model: a hybrid physical model-driven and data-driven approach[J]. IEEE Transactions on Power Systems, 2020, 35 (3): 2475- 2478. DOI
6	YANG Z F, XIE K G, YU J, et al. A general formulation of linear power flow models: basic theory and error analysis[J]. IEEE Transactions on Power Systems, 2019, 34 (2): 1315- 1324. DOI
7	FAN Z X, YANG Z F, YU J, et al. Minimize linearization error of power flow model based on optimal selection of variable space[J]. IEEE Transactions on Power Systems, 2021, 36 (2): 1130- 1140. DOI
8	向明旭, 杨知方, 余娟, 等. 配电网线性潮流模型通式及误差分析[J]. 中国电机工程学报, 2021, 41 (6): 2053- 2064. DOI
	XIANG Mingxu, YANG Zhifang, YU Juan, et al. Linear power flow model in distribution network: unified expression and error analysis[J]. Proceedings of the CSEE, 2021, 41 (6): 2053- 2064. DOI
9	MORELLO R, MUKHOPADHYAY S C, LIU Z, et al. Advances on sensing technologies for smart cities and power grids: a review[J]. IEEE Sensors Journal, 2017, 17 (23): 7596- 7610. DOI
10	鲁鹏, 吕昊, 刘念, 等. 数据-模型混合驱动的配电系统灵活性优化调度[J]. 湘潭大学学报(自然科学版), 2020, 42 (5): 84- 97.
	LU Peng, LV Hao, LIU Nian, et al. Hybrid data and model driven flexibility optimization dispatch of distribution system[J]. Journal of Xiangtan University (Natural Science Edition), 2020, 42 (5): 84- 97.
11	LIU Y X, ZHANG N, WANG Y, et al. Data-driven power flow linearization: a regression approach[J]. IEEE Transactions on Smart Grid, 2019, 10 (3): 2569- 2580. DOI
12	GUO L, ZHANG Y X, LI X L, et al. Data-driven power flow calculation method: a lifting dimension linear regression approach[J]. IEEE Transactions on Power Systems, 2022, 37 (3): 1798- 1808. DOI
13	XU T, WU W C, HONG Y W, et al. Data-driven inverter-based volt/VAr control for partially observable distribution networks[J]. CSEE Journal of Power and Energy Systems, 2021, 9 (2): 548- 560.
14	LIN S, ZHU H. Data-driven modeling for distribution grids under partial observability[C]//2021 North American Power Symposium (NAPS). College Station, TX, USA. IEEE, 2021: 1–6.
15	NOWAK S, CHEN Y C, WANG L W. Measurement-based optimal DER dispatch with a recursively estimated sensitivity model[J]. IEEE Transactions on Power Systems, 2020, 35 (6): 4792- 4802. DOI
16	LI H R, WENG Y, LIAO Y Z, et al. Distribution grid impedance & topology estimation with limited or no micro-PMUs[J]. International Journal of Electrical Power & Energy Systems, 2021, 129, 106794.
17	DEKA D, CHERTKOV M, BACKHAUS S. Joint estimation of topology and injection statistics in distribution grids with missing nodes[J]. IEEE Transactions on Control of Network Systems, 2020, 7 (3): 1391- 1403. DOI
18	陈勇, 李鹏, 张忠军, 等. 基于PCA-GA-LSSVM的输电线路覆冰负荷在线预测模型[J]. 电力系统保护与控制, 2019, 47 (10): 110- 119. DOI
	CHEN Yong, LI Peng, ZHANG Zhongjun, et al. Online prediction model for power transmission line icing load based on PCA-GA-LSSVM[J]. Power System Protection and Control, 2019, 47 (10): 110- 119. DOI
19	陈茜, 李录平, 刘瑞, 等. 基于PCA-KNN融合算法的风力机变桨角度故障诊断方法[J]. 中国电力, 2021, 54 (11): 190- 198. DOI
	CHEN Xi, LI Luping, LIU Rui, et al. Fault diagnosis method of wind turbine pitch angle based on PCA-KNN fusion algorithm[J]. Electric Power, 2021, 54 (11): 190- 198. DOI
20	张立峰, 李晶, 王智. 基于主成分分析和深度神经网络的声学层析成像温度分布重建[J]. 发电技术, 2023, 44 (3): 399- 406. DOI
	ZHANG Lifeng, LI Jing, WANG Zhi. Reconstruction of temperature distribution by acoustic tomography based on principal component analysis and deep nerual network[J]. Power Generation Technology, 2023, 44 (3): 399- 406. DOI
21	LI P H, WU W C, WANG X M, et al. A data-driven linear optimal power flow model for distribution networks[J]. IEEE Transactions on Power Systems, 2023, 38 (1): 956- 959. DOI
22	徐佳宁, 倪裕隆, 朱春波. 基于改进支持向量回归的锂电池剩余寿命预测[J]. 电工技术学报, 2021, 36 (17): 3693- 3704. DOI
	XU Jianing, NI Yulong, ZHU Chunbo. Remaining useful life prediction for lithium-ion batteries based on improved support vector regression[J]. Transactions of China Electrotechnical Society, 2021, 36 (17): 3693- 3704. DOI
23	赵辉, 杨赛, 岳有军, 等. 基于小波分解-卷积神经网络和支持向量回归的短期负荷预测[J]. 科学技术与工程, 2021, 21 (25): 10718- 10724. DOI
	ZHAO Hui, YANG Sai, YUE Youjun, et al. Short-term load forecasting of convolutional neural network support vector regression using wavelet decomposition[J]. Science Technology and Engineering, 2021, 21 (25): 10718- 10724. DOI
24	刘斌, 毕小熊, 党军朋, 等. 基于支持向量回归的变电站蓄电池退化趋势预测[J]. 电源学报, 2020, 18 (6): 207- 214. DOI
	LIU Bin, BI Xiaoxiong, DANG Junpeng, et al. Degradation trend prediction of battery in substation based on support vector regression[J]. Journal of Power Supply, 2020, 18 (6): 207- 214. DOI
25	本文算例数据附录文件[EB/OL](2023-05-21)[2023-07-28].https://www.jianguoyun.com/p/DU1W4wAQjqnoChjdn6AFIAA.
26	吴文传, 张伯明, 巨云涛. 主动配电网网络分析与运行调控[M]. 北京: 科学出版社, 2016: 330-331.

场景	类型	可观测部分	不可观测部分
一	节点	1~15, 19, 20, 23, 26~31	16~18, 21, 22, 24, 25, 32, 33
一	支路	1—2, 2—3, 3—4, 4—5, 5—6, 6—7, 7—8, 8—9, 9—10, 10—11, 11—12, 12—13, 13—14, 14—15, 2—19, 19—20, 3—23, 6—26, 26—27, 27—28, 28—29, 29—30, 30—31	15—16, 16—17, 17—18, 20—21, 21—22, 23—24, 24—25, 31—32, 32—33
二	节点	1~8, 11~15, 19, 20, 23, 26~31	9, 10, 16~18, 21, 22, 24, 25, 32, 33
二	支路	1—2, 2—3, 3—4, 4—5, 5—6, 6—7, 7—8, 10—11, 11—12, 12—13, 13—14, 14—15, 2—19, 19—20, 3—23, 6—26, 26—27, 27—28, 28—29, 29—30, 30—31	15—16, 16—17, 17—18, 20—21, 21—22, 23—24, 24—25, 31—32, 32—33, 8—9, 9—10

场景	类型	可观测部分	不可观测部分
一	节点	1~15, 19, 20, 23, 26~31	16~18, 21, 22, 24, 25, 32, 33
一	支路	1—2, 2—3, 3—4, 4—5, 5—6, 6—7, 7—8, 8—9, 9—10, 10—11, 11—12, 12—13, 13—14, 14—15, 2—19, 19—20, 3—23, 6—26, 26—27, 27—28, 28—29, 29—30, 30—31	15—16, 16—17, 17—18, 20—21, 21—22, 23—24, 24—25, 31—32, 32—33
二	节点	1~8, 11~15, 19, 20, 23, 26~31	9, 10, 16~18, 21, 22, 24, 25, 32, 33
二	支路	1—2, 2—3, 3—4, 4—5, 5—6, 6—7, 7—8, 10—11, 11—12, 12—13, 13—14, 14—15, 2—19, 19—20, 3—23, 6—26, 26—27, 27—28, 28—29, 29—30, 30—31	15—16, 16—17, 17—18, 20—21, 21—22, 23—24, 24—25, 31—32, 32—33, 8—9, 9—10

相对误差	$ \varphi $	场景一		场景二
相对误差	$ \varphi $	平均误差	最大误差	平均误差	最大误差
P_ij	A	1.89×10^–3	6.59×10^–3	1.59×10^–3	7.07×10^–3
	B	1.18×10^–3	5.27×10^–3	2.27×10^–3	1.39×10^–2
	C	1.35×10^–3	7.22×10^–3	1.97×10^–3	8.34×10^–3
Q_ij	A	6.80×10^–3	1.43×10^–2	9.32×10^–3	1.84×10^–2
	B	4.65×10^–3	1.56×10^–2	8.09×10^–3	1.92×10^–2
	C	5.25×10^–3	1.49×10^–2	7.23×10^–3	1.72×10^–2
V	A	3.68×10^–4	8.06×10^–4	4.47×10^–4	9.30×10^–4
	B	3.90×10^–4	9.41×10^–4	5.29×10^–4	1.20×10^–3
	C	4.01×10^–4	8.93×10^–4	5.04×10^–4	1.14×10^–3

相对误差	$ \varphi $	场景一		场景二
相对误差	$ \varphi $	平均误差	最大误差	平均误差	最大误差
P_ij	A	1.89×10^–3	6.59×10^–3	1.59×10^–3	7.07×10^–3
	B	1.18×10^–3	5.27×10^–3	2.27×10^–3	1.39×10^–2
	C	1.35×10^–3	7.22×10^–3	1.97×10^–3	8.34×10^–3
Q_ij	A	6.80×10^–3	1.43×10^–2	9.32×10^–3	1.84×10^–2
	B	4.65×10^–3	1.56×10^–2	8.09×10^–3	1.92×10^–2
	C	5.25×10^–3	1.49×10^–2	7.23×10^–3	1.72×10^–2
V	A	3.68×10^–4	8.06×10^–4	4.47×10^–4	9.30×10^–4
	B	3.90×10^–4	9.41×10^–4	5.29×10^–4	1.20×10^–3
	C	4.01×10^–4	8.93×10^–4	5.04×10^–4	1.14×10^–3

相对误差	$ \varphi $	场景一		场景二
相对误差	$ \varphi $	平均误差	最大误差	平均误差	最大误差
P_ij	A	2.46×10^–3	9.86×10^–3	1.74×10^–3	8.21×10^–3
	B	1.99×10^–3	6.61×10^–3	2.78×10^–3	1.42×10^–2
	C	1.92×10^–3	6.02×10^–3	2.19×10^–3	1.67×10^–2
Q_ij	A	6.40×10^–3	3.28×10^–2	2.66×10^–3	8.94×10^–3
	B	5.54×10^–3	2.90×10^–2	4.28×10^–3	2.06×10^–2
	C	9.85×10^–3	2.41×10^–2	3.80×10^–3	1.82×10^–2
V	A	4.85×10^–4	1.76×10^–3	2.15×10^–4	6.81×10^–4
	B	2.16×10^–4	9.90×10^–4	1.15×10^–4	5.35×10^–4
	C	4.43×10^–4	1.14×10^–3	1.39×10^–4	7.31×10^–4