基于SGMM-MCopula的风光互补系统时空相关场景生成  

Spatiotemporal Correlation Scenario Generation of Wind-Solar Complementary System Based on SGMM-MCopula

作　　者：余意 刘梓轩 赵国汉 刘万 邓友汉 温栋 莫莉[3] YU Yi;LIU Zi-xuan;ZHAO Guo-han;LIU Wan;DENG You-han;WEN Dong;MO Li(Laboratory of Hydro-Wind-Solar Multi-energy Control Coordination,Wuhan 430010,China;China Three Gorges Corporation Science and Technology Research Institute,Beijing 101100,China;School of Civil and Hydraulic Engineering,Huazhong University of Science and Technology,Wuhan 430074,China;China Yangtze Power Co.,Ltd.,Wuhan 430010,China;Three Gorges Jinsha Yunchuan Hydropower Development Co.,Ltd.,Kunming 650204,China)

机构地区：[1]水风光多能互补联合实验室,武汉430010 [2]中国长江三峡集团有限公司科学技术研究院,北京101100 [3]华中科技大学土木与水利工程学院,武汉430074 [4]中国长江电力股份有限公司,武汉430010 [5]三峡金沙江云川水电开发有限公司,昆明650204

出　　处：《科学技术与工程》2025年第10期4156-4167,共12页Science Technology and Engineering

基　　金：中国长江电力股份有限公司资助项目(Z532302054)。

摘　　要：在风能和太阳能大规模并入电网的背景下,电力系统调度策略遭遇了前所未有的挑战。特别是风电和光伏发电的波动性和随机性特征,对系统的稳定性和可控性构成了重大影响。为了精确表征风光电站出力的时空相关特性,并构建具有实际应用价值的场景集,提出了一种基于耦合季节性高斯混合模型(SGMM)与混合Copula函数(MCopula)的风光互补系统时空相关场景生成方法。该方法首先通过构建SGMM捕捉风光出力变量时间序列间的相关特性;其次,采用混合Copula函数来描述变量之间的空间相关特性。在综合时空相关性建模的基础上,结合Copula条件分布函数与逆变换抽样技术,生成了一系列反映时空相关特征的不确定性场景集。仿真实验的结果证实了所提方法的有效性和可靠性,生成的场景集不仅能够较好地反映风光出力的时空相关特征及年内变化趋势,而且与历史实际序列在距离上更为吻合,为电力系统调度提供了强有力的决策支持。研究结果为风光互补系统的不确定性量化提供了新的视角与工具,对于优化电力系统调度策略、降低不确定性风险、促进新能源的高效利用以及推动电力系统的可持续发展,均具有深远的理论和实践意义。In the context of large-scale integration of wind and solar power into the grid,power system dispatch strategies faced unprecedented challenges.The volatility and randomness of wind and photovoltaic power generation significantly impacted system stability and controllability.To accurately characterize the spatiotemporal correlation of wind-solar power output and construct a practically valuable scenario set,a method for generating spatiotemporal correlated scenarios for wind-solar complementary systems was proposed,based on a coupled SGMM(seasonal Gaussian mixture model)and MCopula(mixed Copula function).Initially,the SGMM was constructed to capture the temporal correlation among wind-solar output variables.Then,the mixed Copula function was employed to describe the spatial correlation among variables.Based on the comprehensive modeling of spatiotemporal correlations,a series of uncertainty scenario sets reflecting these characteristics was generated using the Copula conditional distribution function and inverse transform sampling technique.The simulation results confirmed the effectiveness and reliability of the proposed method.The generated scenario sets not only reflected the spatiotemporal correlation characteristics and annual variation trends of wind-solar output but also better matched the historical actual sequences in terms of distance,providing strong decision-making support for power system dispatch.New perspectives and tools were offered for quantifying uncertainties in wind-solar complementary systems,which had profound theoretical and practical significance for optimizing power system dispatch strategies,reducing uncertainty risks,promoting the efficient utilization of renewable energy,and advancing the sustainable development of power systems.

关 键 词：风光出力 时空相关 场景生成 高斯混合模型 COPULA函数 

分 类 号：TM715[电气工程—电力系统及自动化]