【作者】 张靖；

【导师】 舒杰；

【作者基本信息】 中国科学技术大学 ， 动力工程及工程热物理， 2025， 博士

【摘要】 在“双碳”目标的引领下,光伏发电产业已进入快速发展阶段。随着“整县光伏”政策的实施,各地纷纷推动分布式光伏的规模化建设。然而,分布式光伏大规模并网带来了配电网功率波动、线路重过载、电压越限、光伏难消纳等一系列挑战。为降低光伏波动性和间歇性对配电网的负面影响,提高能源利用效率,实现光伏发电的高效消纳,本文从“源-网-系统”三个层面展开研究,旨在为高比例分布式光伏接入的配电网提供一套完整的解决方案。具体开展了以下工作:(1)针对光伏发电功率预测难度大、现有预测方法精度低的问题,提出了一种基于改进门控循环单元(Gate Recurrent Unit,GRU)模型的超短期光伏功率区间预测方法。利用多种方法对GRU模型进行改进,提升了区间预测的可靠性和精度。在晴天相似日和阴雨天相似日的区间预测结果中,本文方法的综合评价指标值分别仅为单一GRU的54.3%和37.4%。(2)针对高比例分布式光伏接入背景下,配电网发生电压越限的问题,提出了一种基于改进差分进化算法的分布式光伏集群优化控制方法。该方法通过集群划分和集群优化控制实现了对配电网电压的调控,避免了电压越限,并促进了分布式光伏的最大化消纳。与传统的集中式优化方法相比,该方法在优化精度上适度妥协,从而显著提高了优化效率,具有更广泛的应用潜力。(3)为了实现智能软开关(Soft Open Point,SOP)在配电网中的最佳配置,实现优化系统潮流和电压分布的目标,提出了一种针对高比例分布式光伏接入的有源配电网柔性互联的多SOP优化配置方法。构建了一种双层优化模型,通过上层的优化配置和下层的优化运行,实现了多配电台区内多个SOP的最优选址与定容。该方法不仅促进了单一配电台区内部的协调,还加强了多个配电台区之间的协同与互补。(4)在SOP优化配置的基础上,为了促进配电网的经济运行,推动“柔性互联”目标的实现,提出了一种基于SOP和多设备协同的含高比例分布式光伏配电网优化调度策略。该策略涵盖了日前优化、日内滚动和实时修正三个阶段,综合考虑了SOP和多设备的协同运行。相比其他策略,该策略不仅能够有效改善配电网的电压波动情况,降低电压越限的风险,还能通过精准的潮流控制实现系统损耗的降低,并提升配电网对光伏的消纳能力。通过对含高比例分布式光伏配电网优化运行技术的深入研究,本文提出了四项关键技术创新。首先,提出的超短期光伏功率区间预测方法有效提升了光伏发电预测的精度和可靠性,为配电网的可靠运行提供了保障;其次,采用的分布式光伏集群优化控制方法有效缓解了配电网电压越限问题,促进了光伏消纳;再次,多SOP优化配置方法提供了一种新的SOP选址定容方案,促进了复杂配电系统的柔性互联;最后,基于SOP与多设备协同的优化调度策略优化了系统潮流与电压分布,降低了电压越限风险,提升了光伏消纳能力。以上研究成果为光伏发电高效消纳和配电网优化运行提供了重要理论支持和工程应用指导。更多还原

【Abstract】 Guided by the goals of achieving“carbon peaking and carbon neutrality”,the photovoltaic(PV)power generation industry has embarked on a phase of accelerated growth.With the implementation of the“whole county PV”policy,various regions have actively promoted the large-scale construction of distributed PVs.However,large-scale grid integration of distributed PVs has brought about a series of challenges,including power fluctuations in the distribution network,line overloads,voltage violations,and difficulties in PV consumption.To mitigate the adverse effects of PV volatility and intermittency on distribution networks,enhance energy utilization efficiency,and achieve efficient PV power integration,this thesis undertakes research from the“source-grid-system”perspective,aiming to provide a comprehensive solution for distribution networks with high-penetration distributed PVs.The specific tasks carried out in this research are as follows:(1)To address the difficulties in PV power prediction and the limitations in accuracy of current methodologies,an ultra-short-term PV power interval prediction approach based on an improved gate recurrent unit(GRU)model is proposed.Various methods are used to improve the GRU model,enhancing the reliability and accuracy of the interval prediction.In the interval prediction results for sunny and rainy similar days,the comprehensive evaluation indicator values of the proposed method are only 54.3%and 37.4%,respectively,compared to a single GRU model.(2)In the context of high-penetration distributed PV integration,voltage violations in the distribution network are a significant issue.An optimal control method for distributed PV clusters based on an improved differential evolution algorithm is proposed.This method achieves voltage regulation in the distribution network through cluster partitioning and optimal control,preventing voltage violations and promoting the maximum consumption of distributed PV.Compared to traditional centralized optimization methods,this approach makes a moderate trade-off in optimization accuracy to significantly improve optimization efficiency,demonstrating broader application potential.(3)To achieve the optimal configuration of soft open point(SOP)in the distribution network and optimize the system’s power flow and voltage distribution,a multi-SOP optimal configuration method for active distribution networks with high-penetration distributed PV is proposed.A bi-level optimization framework is established,where the upper-level optimal configuration and the lower-level optimal operation jointly realize the optimal siting and sizing of multiple SOPs across multiple distribution stations.This method not only facilitates coordination within individual distribution stations but also strengthens coordination and complementarity between multiple distribution stations.(4)Based on the optimal configuration of SOPs,an optimal dispatching strategy for distribution networks with high-penetration distributed PV,integrating SOPs and multi-device coordination,is proposed to promote the economic operation of the distribution networks and facilitate the achievement of“flexible interconnection”goals.This strategy encompasses day-ahead optimization,intra-day rolling,and real-time correction,considering the coordinated operation of SOPs and multiple devices.Compared to other strategies,this approach effectively improves voltage fluctuations in the distribution network,reduces the risk of voltage violations,reduces system losses through precise power flow control,and enhances the distribution network’s ability to absorb PV power.Through in-depth research on the optimal operation technology of distribution networks with high-penetration distributed PV,this thesis presents four key technological innovations.First,the proposed ultra-short-term PV power interval prediction method significantly improves the accuracy and reliability of PV generation predictions,thereby securing the dependable operation of the distribution network.Second,the distributed PV cluster optimal control method effectively alleviates the issue of voltage violations in the distribution network,promoting the consumption of PV generation.Third,the multi-SOP optimal configuration method offers a novel method for the siting and sizing of SOPs,facilitating the flexible interconnection of complex distribution systems.Lastly,the optimal dispatching strategy based on the collaboration between SOPs and multiple devices optimizes system power flow and voltage distribution,reduces the risk of voltage violations,and enhances the ability to accommodate PV generation.The aforementioned research results provide essential theoretical support and engineering guidance for the efficient integration of PV generation and the optimal operation of distribution networks.更多还原