小样本数据驱动模式下的新建微电网优化调度策略

doi:10.16183/j.cnki.jsjtu.2023.394

摘要/Abstract

摘要：

新建微电网缺少历史运行数据,常规数据驱动的方法难以精确预测可再生能源出力,进而影响调度计划制定的准确性.为此,提出一种适用于新建微电网小样本数据场景的微电网优化调度方法.首先,设计融合域对抗神经网络和长短期记忆网络的改进网络结构,将域对抗思想和梯度反转机制引入迁移学习中,提高模型泛化能力,减小数据的域分布差异,使用出力特征相似电站的丰富运行数据对目标电站出力进行预测,克服小样本条件下出力预测精度不高的问题.进一步,将优化调度模型转化为马尔可夫决策过程,使用双延迟深度确定性策略梯度算法求解.最后,以改进CIGRE 14节点微电网为例验证了所提方法的有效性.

关键词: 小样本, 可再生能源出力, 对抗迁移学习, 深度强化学习, 微电网优化调度

Abstract:

Newly built microgrids lack historical operation data, making it challenging to predict renewable power output accurately using conventional data-driven methods, which in turn affects the accuracy of scheduling plans. To address this problem, an optimal scheduling method for newly built microgrids in scenarios with limited sample data is proposed. First, an improved network structure integrating a domain adversarial neural network with a long-short-term memory network is designed. The domain adversarial approach and gradient inversion mechanism are incorporated into transfer learning to improve the generalization ability of the model. This reduces the domain distribution discrepancy in the data, and uses the rich operation data of power stations with similar output characteristics to predict the output of the target station, which overcomes the challenge of poor accuracy under the conditions of small samples. Additionally, the optimal scheduling model is transformed into a Markov decision process and solved using double-delay deep deterministic policy gradient algorithm. Finally, the effectiveness of the proposed method is validated through a case study involving an improved CIGRE 14-node microgrid.

Key words: small sample, renewable energy contribution, adversarial transfer learning, deep reinforcement learning, optimal scheduling of microgrid

中图分类号:

TM732

陈实, 杨林森, 刘艺洪, 罗欢, 臧天磊, 周步祥. 小样本数据驱动模式下的新建微电网优化调度策略[J]. 上海交通大学学报, 2025, 59(6): 732-745.

CHEN Shi, YANG Linsen, LIU Yihong, LUO Huan, ZANG Tianlei, ZHOU Buxiang. Optimal Scheduling Strategy of Newly-Built Microgrid in Small Sample Data-Driven Mode[J]. Journal of Shanghai Jiao Tong University, 2025, 59(6): 732-745.

图/表 12

图1

图2

图3

表1

表2

图4

图5

图6

图7

图8

图9

图10

参考文献 22

[22]	LIANG Hong, LI Hongxin, ZHANG Huaying, et al. Control strategy of microgrid energy storage system based on deep reinforcement learning[J]. Power System Technology, 2021, 45(10): 3869-3877.
[23]	黎海涛, 申保晨, 杨艳红, 等. 基于改进竞争深度Q网络算法的微电网能量管理与优化策略[J]. 电力系统自动化, 2022, 46(7): 42-49.
	LI Haitao, SHEN Baochen, YANG Yanhong, et al. Energy management and optimization strategy for microgrid based on improved dueling deep Q network algorithm[J]. Automation of Electric Power Systems, 2022, 46(7): 42-49.
[24]	冯昌森, 张瑜, 文福拴, 等. 基于深度期望Q网络算法的微电网能量管理策略[J]. 电力系统自动化, 2022, 46(3): 14-22.
	FENG Changsen, ZHANG Yu, WEN Fushuan, et al. Energy management strategy for microgrid based on deep expected Q network algorithm[J]. Automation of Electric Power Systems, 2022, 46(3): 14-22.
[25]	叶宇剑, 袁泉, 汤奕, 等. 抑制柔性负荷过响应的微网分散式调控参数优化[J]. 中国电机工程学报, 2022, 42(5): 1748-1759.
	YE Yujian, YUAN Quan, TANG Yi, et al. Decentralized coordination parameters optimization in microgrids mitigating demand response synchronization effect of flexible loads[J]. Proceedings of the CSEE, 2022, 42(5): 1748-1759.
[26]	赵鹏杰, 吴俊勇, 王燚, 等. 基于深度强化学习的微电网优化运行策略[J]. 电力自动化设备, 2022, 42(11): 9-16.
	ZHAO Pengjie, WU Junyong, WANG Yi, et al. Optimal operation strategy of microgrid based on deep reinforcement learning[J]. Electric Power Automation Equipment, 2022, 42(11): 9-16.
[27]	GANIN Y, USTINOVA E, AJAKAN H, et al. Domain-adversarial training of neural networks[J]. The Journal of Machine Learning Research, 2016, 17(1): 2096-2030.
[28]	邵振国, 张承圣, 陈飞雄, 等. 生成对抗网络及其在电力系统中的应用综述[J]. 中国电机工程学报, 2023, 43(3): 987-1004.
	SHAO Zhenguo, ZHANG Chengsheng, CHEN Fei-xiong, et al. A review on generative adversarial networks for power system applications[J]. Proceedings of the CSEE, 2023, 43(3): 987-1004.
[29]	史加荣, 王丹, 尚凡华, 等. 随机梯度下降算法研究进展[J]. 自动化学报, 2021, 47(9): 2103-2119.
	SHI Jiarong, WANG Dan, SHANG Fanhua, et al. Research advances on stochastic gradient descent algorithms[J]. Acta Automatica Sinica, 2021, 47(9): 2103-2119.
[30]	何颖源, 陈永翀, 刘勇, 等. 储能的度电成本和里程成本分析[J]. 电工电能新技术, 2019, 38(9): 1-10. doi: 10.12067/ATEEE1907045
	HE Yingyuan, CHEN Yongchong, LIU Yong, et al. Analysis of cost per kilowatt-hour and cost per mileage for energy storage technologies[J]. Advanced Technology of Electrical Engineering & Energy, 2019, 38(9): 1-10.
[31]	冯斌, 胡轶婕, 黄刚, 等. 基于深度强化学习的新型电力系统调度优化方法综述[J]. 电力系统自动化, 2023, 47(17): 187-199.
	FENG Bin, HU Yijie, HUANG Gang, et al. Review on optimization methods for new power system dispatch based on deep reinforcement learning[J]. Automation of Electric Power Systems, 2023, 47(17): 187-199.
[32]	FUJIMOTO S, VAN HOOF H, MEGER D. Addressing function approximation error in actor-critic methods[C]// Proceedings of the 35th International Conference on Machine Learning. Stockholm, Sweden: PMLR, 2018: 1587-1596.
[33]	PAPATHANASSIOU S, HATZIARGYRIOU N, STRUNZ K. A benchmark low voltage microgrid network[C]// Proceedings of the CIGRE Symposium: Power Systems with Dispersed Generation. Athens, Greece: CIGRE, 2005: 1-8.
[34]	PAN S J, YANG Q. A survey on transfer learning[J]. IEEE Transactions on Knowledge & Data Engineering, 2010, 22(10): 1345-1359.
[35]	CCE. California ISO open access same-time information system (OASIS)[DB/OL]. (2022-12-02)[2023-07-28]. https://www.energyfreedomco.org/elec-system.php.
[1]	王文彬, 郑蜀江, 范瑞祥, 等. “双碳” 背景下微网分布式电能交易绩效评价指标与方法[J]. 上海交通大学学报, 2022, 56(3): 312-324. doi: 10.16183/j.cnki.jsjtu.2021.391
	WANG Wenbin, ZHENG Shujiang, FAN Ruixiang, et al. Performance evaluation index and method of micro-grid distributed electricity trading under the background of “carbon peaking and carbon neutrality”[J]. Journal of Shanghai Jiao Tong University, 2022, 56(3): 312-324.
[2]	ASLAM S, HERODOTOU H, MOHSIN S M, et al. A survey on deep learning methods for power load and renewable energy forecasting in smart microgrids[J]. Renewable & Sustainable Energy Reviews, 2021, 144: 110992.
[3]	杨茂, 王金鑫. 考虑可再生能源出力不确定的孤岛型微电网优化调度[J]. 中国电机工程学报, 2021, 41(3): 973-985.
	YANG Mao, WANG Jinxin. Optimal scheduling of islanded microgrid considering uncertain output of renewable energy[J]. Proceedings of the CSEE, 2021, 41(3): 973-985.
[4]	GAMIL M M, UEDA S, NAKADOMARI A, et al. Optimal multi-objective power scheduling of a residential microgrid considering renewable sources and demand response technique[J]. Sustainability, 2022, 14(21): 13709.
[5]	WANG J H, YAN G W, REN M F, et al. Short term photovoltaic power prediction based on transfer learning and considering sequence uncertainty[J]. Journal of Renewable & Sustainable Energy, 2023, 15(1): 013501.
[6]	乔颖, 孙荣富, 丁然, 等. 基于数据增强的分布式光伏电站群短期功率预测(一): 方法框架与数据增强[J]. 电网技术, 2021, 45(5): 1799-1808.
	QIAO Ying, SUN Rongfu, DING Ran, et al. Distributed photovoltaic station cluster gridding short-term power forecasting part I: Methodology and data augmentation[J]. Power System Technology, 2021, 45(5): 1799-1808.
[7]	唱友义, 孙赫阳, 顾泰宇, 等. 采用历史数据扩充方法的风力发电量月度预测[J]. 电网技术, 2021, 45(3): 1059-1068.
	CHANG Youyi, SUN Heyang, GU Taiyu, et al. Monthly forecast of wind power generation using historical data expansion method[J]. Power System Technology, 2021, 45(3): 1059-1068.
[8]	LIU G L, ZHANG S W, ZHAO H, et al. Super-resolution perception for wind power forecasting by enhancing historical data[J]. Frontiers in Energy Research, 2022, 10: 959333.
[9]	史凯钰, 张东霞, 韩肖清, 等. 基于LSTM与迁移学习的光伏发电功率预测数字孪生模型[J]. 电网技术, 2022, 46(4): 1363-1372.
	SHI Kaiyu, ZHANG Dongxia, HAN Xiaoqing, et al. Digital twin model of photovoltaic power generation prediction based on LSTM and transfer learning[J]. Power System Technology, 2022, 46(4): 1363-1372.
[10]	魏泽涛, 刘友波, 沈晓东, 等. 基于样本数据迁移学习的贫资料地区小水电超短期出力建模及发电预测[J]. 中国电机工程学报, 2023, 43(7): 2652-2666.
	WEI Zetao, LIU Youbo, SHEN Xiaodong, et al. Ultra-short-term power generation modeling and prediction for small hydropower in data-scarce areas based on sample data transfer learning[J]. Proceedings of the CSEE, 2023, 43(7): 2652-2666.
[11]	LUO X, ZHANG D X, ZHU X. Combining transfer learning and constrained long short-term memory for power generation forecasting of newly-constructed photovoltaic plants[J]. Renewable Energy, 2022, 185: 1062-1077.
[12]	程凯, 彭小圣, 徐其友, 等. 基于特征选择与多层级深度迁移学习的风电场短期功率预测[J]. 高电压技术, 2022, 48(2): 497-503.
	CHENG Kai, PENG Xiaosheng, XU Qiyou, et al. Short-term wind power prediction based on feature selection and multi-level deep transfer learning[J]. High Voltage Engineering, 2022, 48(2): 497-503.
[13]	张童彦, 廖清芬, 唐飞, 等. 基于气象资源插值与迁移学习的广域分布式光伏功率预测方法[J]. 中国电机工程学报, 2023, 43(20): 7929-7940.
	ZHANG Tongyan, LIAO Qingfen, TANG Fei, et al. Wide-area distributed photovoltaic power forecast method based on meteorological resource interpolation and transfer learning[J]. Proceedings of the CSEE, 2023, 43(20): 7929-7940.
[14]	米阳, 彭建伟, 陈博洋, 等. 基于一致性原理和梯度下降法的微电网完全分布式优化调度[J]. 电力系统保护与控制, 2022, 50(15): 1-10.
	MI Yang, PENG Jianwei, CHEN Boyang, et al. Fully distributed optimal dispatch of a microgrid based on consensus principle and gradient descent[J]. Power System Protection & Control, 2022, 50(15): 1-10.
[15]	陆秋瑜, 于珍, 杨银国, 等. 考虑源荷功率不确定性的海上风力发电多微网两阶段优化调度[J]. 上海交通大学学报, 2022, 56(10): 1308-1316. doi: 10.16183/j.cnki.jsjtu.2021.409
	LU Qiuyu, YU Zhen, YANG Yinguo, et al. Two-stage optimal schedule of offshore wind-power-integrated multi-microgrid considering uncertain power of sources and loads[J]. Journal of Shanghai Jiao Tong University, 2022, 56(10): 1308-1316.
[16]	夏超英, 苗海丽. 基于二次型最优控制的微电网实时能量管理策略[J]. 中国电机工程学报, 2019, 39(3): 721-730.
	XIA Chaoying, MIAO Haili. Real-time energy management strategy for micro-grid based on the quadratic optimal control theory[J]. Proceedings of the CSEE, 2019, 39(3): 721-730.
[17]	姚建国, 余涛, 杨胜春, 等. 提升电网调度中人工智能可用性的混合增强智能知识演化技术[J]. 电力系统自动化, 2022, 46(20): 1-12.
	YAO Jianguo, YU Tao, YANG Shengchun, et al. Knowledge evolution technology based on hybrid-augmented intelligence for improving practicability of artificial intelligence in power grid dispatch[J]. Automation of Electric Power Systems, 2022, 46(20): 1-12.
[18]	陈亭轩, 徐潇源, 严正, 等. 基于深度强化学习的光储充电站储能系统优化运行[J]. 电力自动化设备, 2021, 41(10): 90-98.
	CHEN Tingxuan, XU Xiaoyuan, YAN Zheng, et al. Optimal operation based on deep reinforcement learning for energy storage system in photovoltaic-storage charging station[J]. Electric Power Automation Equipment, 2021, 41(10): 90-98.
[19]	杨挺, 赵黎媛, 王成山. 人工智能在电力系统及综合能源系统中的应用综述[J]. 电力系统自动化, 2019, 43(1): 2-14.
	YANG Ting, ZHAO Liyuan, WANG Chengshan. Review on application of artificial intelligence in power system and integrated energy system[J]. Automation of Electric Power Systems, 2019, 43(1): 2-14.
[20]	吴倩红, 韩蓓, 冯琳, 等. “人工智能+” 时代下的智能电网预测分析[J]. 上海交通大学学报, 2018, 52(10): 1206-1219. doi: 10.16183/j.cnki.jsjtu.2018.10.008
	WU Qianhong, HAN Bei, FENG Lin, et al. “AI+” based smart grid prediction analysis[J]. Journal of Shanghai Jiao Tong University, 2018, 52(10): 1206-1219.
[21]	CHEN S, LIU Y H, GUO Z W, et al. Deep reinforcement learning based research on low-carbon scheduling with distribution network schedulable resources[J]. IET Generation, Transmission & Distribution, 2023, 17(10): 2289-2300.
[22]	梁宏, 李鸿鑫, 张华赢, 等. 基于深度强化学习的微网储能系统控制策略研究[J]. 电网技术, 2021, 45(10): 3869-3877.

电站编号	经度/(°)	纬度/(°)
1号源域	-86.9124	36.8796
2号源域	-87.0241	37.1012
3号源域	-86.8248	36.4638
4号源域	-86.7568	36.5879
1号光伏	-86.9442	36.9118
2号光伏	-86.9365	36.9158
1号风电站	-86.9432	36.9089
2号风电站	-86.9412	36.9093

出力类型	相关系数
出力类型	温度	空气密度	辐射强度	风速	降水量
WT	-0.22	0.16	-0.23	0.97	0.04
PV	0.39	0.22	0.95	-0.21	-0.06