基于滚动时域优化和反馈校正的虚拟电厂实时优化调度方法  

A Real-Time Optimal Scheduling Method of Virtual Power Plants Based on Rolling Time Domain Optimization and Feedback Correction

作　　者：蒋望 赵理 李亦非[2] 张同尊 于钊 彭静 JIANG Wang;ZHAO Li;LI Yifei;ZHANG Tongzun;YU Zhao;PENG Jing(State Grid Chongqing Electric Power Company,Chongqing 400015,China;State Grid Beijing Electric Power Company,Beijing 100075,China;State Grid Shandong Linyi Power Supply Company,Linyi 276000,Shandong,China)

机构地区：[1]国网重庆市电力公司,重庆400015 [2]国网北京市电力公司,北京100075 [3]国网山东临沂供电公司,山东临沂276000

出　　处：《电网与清洁能源》2022年第8期40-50,共11页Power System and Clean Energy

基　　金：国家自然科学基金项目(52061130217)。

摘　　要：虚拟电厂(virtual power plants,VPPs)已经成为对分布式新能源进行调度控制的重要方式,而由于日前新能源预测和负荷预测的误差,经常会影响调度策略并造成VPPs次优运行。提出了一种基于反馈校正控制模型预测控制的VPP自适应预测优化调度策略,以实现不确定性条件下VPPs的最优运行。该策略包括2个部分:滚动时域优化(receding horizon optimization,RHO)和反馈校正(feedback correction,FC)。RHO采用时间序列模型与卡尔曼滤波相结合的混合预测算法对RES和负载的输出功率进行预测,并根据预测结果进行调度。FC采用基于快速滚动灰色模型的超短期误差预测,对RHO策略进行调整。FC用于最小化调整,以补偿预测误差。所提策略在某实际配电系统的VPP上实施,结果表明,其能更好地跟踪系统内实际可用资源,且供需之间的不匹配程度最小。The virtual power plants(VPP)has become an important development trend of distributed energy industry due to its efficient management and control of distributed energy.Most of the operational strategies of VPPs are based on the prior day forecast data,and the forecast errors of renewable energy sources(RES)and loads can seriously affect the power dispatch strategy,resulting in sub-optimal operation of VPPs. To this end,this paper proposes a VPP adaptive predictive optimization scheduling strategy based on feedback correction model predictive control to achieve optimal operation of VPPs under uncertain conditions. The strategy includes two parts: Receding Horizon Optimization(RHO) and feedback correction(FC).RHO uses a hybrid prediction algorithm combining time series model and Kalman filter to predict the output power of RES and load,and then schedules according to the prediction results. FC uses ultra-short-term error prediction based on fast rolling gray model to adjust the RHO strategy. FC is used to minimize adjustments to compensate for prediction errors. The proposed strategy is implemented on the VPP of an actual distribution system,and the simulation results prove the effectiveness of the proposed method,that is,it can better track the actual available resources in the system,and the mismatch between supply and demand is minimal.

关 键 词：虚拟电厂 分布式能源资源 模型预测控制 能源管理 

分 类 号：TM28[一般工业技术—材料科学与工程]