End-to-End Collaborative Optimization Method for Microgrid Power Prediction and Optimal Scheduling

doi:10.16183/j.cnki.jsjtu.2024.224

Abstract

Abstract:

Microgrids, as one of the effective methods for integrating new energy sources, play a crucial role in the new-type power systems. In microgrids with high renewable energy penetration, the objectives of renewable energy power forecasting and microgrid optimal scheduling may be misaligned. To address this issue, this study proposes an end-to-end optimization model which combines power forecasting with day-ahead and intraday scheduling to maximize the operational benefits of the microgrid. It also provides a corresponding solution method. Initially, a bi-level optimization framework is established. The upper level focuses on training the power forecasting model, formulated as a combined forecasting problem, while the lower level aims to minimize microgrid operational costs. The result of the lower-level optimization is used as the loss function to optimize the forecasting weights in the upper level. Subsequently, a heuristic algorithm iteratively is employed to solve the upper and lower level problems, thereby obtaining forecasting results and scheduling schemes which minimize the operational costs. Finally, the effectiveness of the proposed method in enhancing microgrid operational benefits is validated by integrating real renewable energy data into a typical microgrid extended from the IEEE 33-node and IEEE 123-node systems.

Key words: microgrid scheduling, new energy prediction, composite forecasting, end-to-end optimization

CLC Number:

TM732

ZHANG Li, WANG Bao, JIA Jianxiong, SONG Zhumeng, YE Yutong, YU Yue, LIN Jiaqing, XU Xiaoyuan. End-to-End Collaborative Optimization Method for Microgrid Power Prediction and Optimal Scheduling[J]. Journal of Shanghai Jiao Tong University, 2025, 59(6): 720-731.

Figures/Tables 16

Fig.1

Tab.1

Tab.2

Tab.3

Tab.4

Tab.5

Fig.2

Fig.3

Tab.6

Fig.4

Fig.5

Fig.6

Fig.7

Fig.8

Tab.7

Fig.9

References 33

[1]	韩肖清, 李廷钧, 张东霞, 等. 双碳目标下的新型电力系统规划新问题及关键技术[J]. 高电压技术, 2021, 47(9): 3036-3046.
	HAN Xiaoqing, LI Tingjun, ZHANG Dongxia, et al. New issues and key technologies of new power system planning under double carbon goals[J]. High Voltage Engineering, 2021, 47(9): 3036-3046.
[2]	郑漳华, 艾芊. 微电网的研究现状及在我国的应用前景[J]. 电网技术, 2008, 32(16): 27-31.
	ZHENG Zhanghua, AI Qian. Present situation of research on microgrid and its application prospects in China[J]. Power System Technology, 2008, 32(16): 27-31.
[3]	杨新法, 苏剑, 吕志鹏, 等. 微电网技术综述[J]. 中国电机工程学报, 2014, 34(1): 57-70.
	YANG Xinfa, SU Jian, LÜ Zhipeng, et al. Overview on micro-grid technology[J]. Proceedings of the CSEE, 2014, 34(1): 57-70.
[4]	FAN H, LI Y L, YU B, et al. Stochastic optimal source-load-storage coordinated dispatch in microgird[C]// 2020 International Conference on Smart Grids and Energy Systems. Perth, Australia: IEEE, 2020: 746-750.
[5]	PEKASLAN D, WAGNER C, GARIBALDI J M, et al. Uncertainty-aware forecasting of renewable energy sources[C]// 2020 IEEE International Conference on Big Data and Smart Computing. Busan, Korea: IEEE, 2020: 240-246.
[6]	刘一欣, 郭力, 王成山. 微电网两阶段鲁棒优化经济调度方法[J]. 中国电机工程学报, 2018, 38(14): 4013-4022.
	LIU Yixin, GUO Li, WANG Chengshan. Economic dispatch of microgrid based on two stage robust optimization[J]. Proceedings of the CSEE, 2018, 38(14): 4013-4022.
[7]	FAN H Z, ZHANG G R, ZHANG Y, et al. Multiple time-scale optimization scheduling for microgrids[C]// 2018 Chinese Automation Congress. Xi’an, China: IEEE, 2018: 3544-3549.
[8]	KLOUBERT M L, SCHWIPPE J, MÜLLER S C, et al. Analyzing the impact of forecasting errors on redispatch and control reserve activation in congested transmission networks[C]// 2015 IEEE Eindhoven PowerTech. Eindhoven, Netherlands: IEEE, 2015: 1-6.
[9]	MORALES J M, MUÑOZ M A, PINEDA S. Prescribing net demand for two-stage electricity generation scheduling[J]. Operations Research Perspectives, 2023, 10: 100268.
[10]	SANG L W, XU Y L, LONG H, et al. Safety-aware semi-end-to-end coordinated decision model for voltage regulation in active distribution network[J]. IEEE Transactions on Smart Grid, 2023, 14(3): 1814-1826.
[11]	WAHDANY D, SCHMITT C, CREMER J L. More than accuracy: End-to-end wind power forecasting that optimises the energy system[J]. Electric Power Systems Research, 2023, 221: 109384.
[12]	CHEN X B, YANG Y F, LIU Y K, et al. Feature-driven economic improvement for network-constrained unit commitment: A closed-loop predict-and-optimize framework[J]. IEEE Transactions on Power Systems, 2022, 37(4): 3104-3118.
[13]	FISHER M L. The Lagrangian relaxation method for solving integer programming problems[J]. Management Science, 1981, 27(1): 1-18.
[14]	LI G, CHIANG H D. Toward cost-oriented forecasting of wind power generation[J]. IEEE Transactions on Smart Grid, 2018, 9(4): 2508-2517.
[15]	ZHOU X, WANG J Y, WANG X Y, et al. Deep reinforcement learning for microgrid operation optimization: A review[C]// 2023 8th Asia Conference on Power and Electrical Engineering, Tianjin, China: IEEE, 2023: 2059-2065.
[16]	CHAKRABORTY S, SIMOES M G. PV-microgrid operational cost minimization by neural forecasting and heuristic optimization[C]// 2008 IEEE Industry Applications Society Annual Meeting. Edmonton, Canada: IEEE, 2008: 1-8.
[17]	MU C X, SHI Y K, XU N, et al. Multi-objective interval optimization dispatch of microgrid via deep reinforcement learning[J]. IEEE Transactions on Smart Grid, 2024, 15(3): 2957-2970.
[18]	BARREDO ARRIETA A, DÍAZ-RODRÍGUEZ N, DEL SER J, et al. Explainable artificial intelligence (XAI): Concepts, taxonomies, opportunities and challenges toward responsible AI[J]. Information Fusion, 2020, 58: 82-115.
[19]	王成山, 武震, 李鹏. 微电网关键技术研究[J]. 电工技术学报, 2014, 29(2): 1-12.
	WANG Chengshan, WU Zhen, LI Peng. Research on key technologies of microgrid[J]. Transactions of China Electrotechnical Society, 2014, 29(2): 1-12.
[20]	LI L, XU S Y. Optimal day-ahead scheduling of microgrid participating in energy and spinning reserve markets[C]// 2020 5th Asia Conference on Power and Electrical Engineering, Chengdu, China: IEEE, 2020: 1049-1055.
[21]	FARIVAR M, LOW S H. Branch flow model: Relaxations and convexification: Part I[J]. IEEE Transactions on Power Systems, 2013, 28(3): 2554-2564.
[22]	XIAO L, WANG J Z, DONG Y, et al. Combined forecasting models for wind energy forecasting: A case study in China[J]. Renewable & Sustainable Energy Reviews, 2015, 44: 271-288.
[23]	张国强, 张伯明. 基于组合预测的风电场风速及风电机功率预测[J]. 电力系统自动化, 2009, 33(18): 92-95.
	ZHANG Guoqiang, ZHANG Boming. Wind speed and wind turbine output forecast based on combination method[J]. Automation of Electric Power Systems, 2009, 33(18): 92-95.
[24]	WANG D S, TAN D P, LIU L. Particle swarm optimization algorithm: An overview[J]. Soft Computing, 2018, 22(2): 387-408.
[25]	KENNEDY J, EBERHART R. Particle swarm optimization[C]// Proceedings of ICNN’95-International Conference on Neural Networks. Perth, Australia: IEEE, 1995: 1942-1948.
[26]	AWAD M, KHANNA R. Support vector regression[M]//Efficient learning machines. Berkeley, USA: Apress, 2015: 67-80.
[27]	POPESCU M-C, BALAS V E, PERESCU-POPESCU L, et al. Multilayer perceptron and neural networks[J]. WSEAS Transactions on Circuits & Systems, 2009, 8(7): 579-588.
[28]	SEGAL M R. Machine learning benchmarks and random forest regression[DB/OL]. (2003-04-14)[2024-06-12]. https://escholarship.org/uc/item/35x3v9t4.
[29]	GURYANOV A. Histogram-based algorithm for building gradient boosting ensembles of piecewise linear decision trees[M]//Lecture notes in computer science. Cham: Springer, 2019: 39-50.
[30]	DOLATABADI S H, GHORBANIAN M, SIANO P, et al. An enhanced IEEE 33 bus benchmark test system for distribution system studies[J]. IEEE Transactions on Power Systems, 2021, 36(3): 2565-2572.
[31]	陈池瑶, 苗世洪, 姚福星, 等. 基于多智能体算法的多微电网-配电网分层协同调度策略[J]. 电力系统自动化, 2023, 47(10): 57-65.
	CHEN Chiyao, MIAO Shihong, YAO Fuxing, et al. Hierarchical cooperative dispatching strategy of multi-microgrid and distribution networks based on multi-agent algorithm[J]. Automation of Electric Power Systems, 2023, 47(10): 57-65.
[32]	HONG T, PINSON P, FAN S, et al. Probabilistic energy forecasting: Global energy forecasting competition 2014 and beyond[J]. International Journal of Forecasting, 2016, 32(3): 896-913.
[33]	CHAI Y Y, GUO L, WANG C S, et al. Network partition and voltage coordination control for distribution networks with high penetration of distributed PV units[J]. IEEE Transactions on Power Systems, 2018, 33(3): 3396-3407.

模型	参数
SVR	kernel为poly, C=0.5, γ=0.03
MLP	hidden_layer_size为75, max_iter为30
RFR	n_estimators为5, max_depth为5, min_samples_leaf为7
HGB	max_iter为20, max_leaf_nodes为20, min_samples_leaf为7

分布式电源类型	参数	取值
燃气轮机1,2	最大容量/kW	600
	向上爬坡速率/(kW·h^-1)	125
	向下爬坡速率/(kW·h^-1)	125
	成本系数,a/(元·kW^-2)	0.0002
	成本系数,b/(元·kW^-2)	0.386
	成本系数,c/元	0
光伏电站1	最大容量/(kV·A)	2000
光伏电站2	最大容量/(kV·A)	2400
光伏电站3	最大容量/(kV·A)	1800
储能设备1,2	电池容量/(kW·h)	1600
	最大充放电功率/kW	500
	充放电效率	0.95
	荷电状态下限	0.1
	荷电状态上限	0.95

对象	参数	取值
燃气轮机	向上再调度成本	0.412
	向下再调度成本	0.386
电力交易	购电量增加的成本	1.200
	购电量减少的成本	0.400
	售电量增加的成本	0.280
	售电量减少的成本	0.400

时段	购电价格	售电价格
峰时段(8:00—11:00, 13:00—16:00, 18:00—22:00)	1.189	0.352
谷时段(0:00—7:00, 22:00—24:00)	0.423	0.352
平时段(7:00—8:00, 11:00—13:00, 16:00—18:00)	0.738	0.352

模型	e_MSE	e_MAE	R²
SVR	0.00978	0.0482	0.8243
MLP	0.00984	0.0672	0.8410
RFR	0.00889	0.0605	0.7979
HGB	0.00927	0.0622	0.8504
精度导向组合预测	0.00764	0.0487	0.8497
收益导向组合预测	0.00931	0.0568	0.8025