电工技术学报2025，Vol.40Issue(15)：4755-4769,15.DOI:10.19595/j.cnki.1000-6753.tces.241189
计及电力-通信-交通耦合网络不确定性的虚拟电厂鲁棒优化调度
Robust Optimization Scheduling of Virtual Power Plant Considering the Uncertainty of Power-Communication-Transportation Coupling Network
摘要
Abstract
As innovations in intersecting technologies,such as renewable energy generation and electrified transportation,continue to progress,the connections among the power,communication,and transportation systems are becoming increasingly closer.In the coupling network of the power-communication-transportation system,virtual power plant(VPP)utilizes advanced control,communication,and information collection technologies to aggregate and intelligently control distributed resources,actively responding to grid demands,thereby enhancing grid operational stability.However,existing VPP optimization scheduling methods overlook the impact of uncertainties in the power-communication-transportation coupling network on the optimization scheduling of VPP demand response,leading to high costs and poor robustness of scheduling.Therefore,researching a robust optimization scheduling method for VPP demand response that considers the uncertainties in the power-communication-transportation coupling network is of great significance for enhancing system reliability,security,and economic efficiency.To address this issue,this paper proposes a robust optimization scheduling method of VPP considering the uncertainty of the power-communication-transportation coupling network. Firstly,a model of the coupling power-communication-transportation system is constructed.In the power-communication-transportation coupling model,the power grid side achieves the aggregation of resources such as renewable energy,energy storage,controllable loads,and EVs through the VPP,coordinating these resources to participate in demand response.On the transportation network side,fast-charging stations facilitate coupling with the power grid by providing charging services for EVs and allowing aggregated EV resources to participate in VPP control through EV aggregators.On the communication network side,the VPP interacts with the distribution network control center and its associated resources,exchanging parameter information and issuing control commands via heterogeneous communication methods such as Ethernet,power line carrier,and wireless networks.The optimization problem is formulated with the objective of minimizing the weighted sum of grid loss,node voltage deviation,and VPP economic cost. Subsequently,uncertainties from the three networks are analyzed.The main uncertainty factors on the communication network side include the operational reliability of control terminals,the effectiveness of communication bandwidth fluctuations,communication delay in control commands,and communication quality.Uncertainty factors on the power grid originate from the variability in renewable energy output.This paper models power grid side uncertainty using the example of distributed photovoltaic output uncertainty.The transportation network side uncertainty mainly stems from the unpredictability of EV demand response,as the response level of privately owned EVs to incentive pricing fluctuates within a certain range.Based on the uncertainty model,a robust optimization scheduling problem considering uncertainties in the power-communication-transportation network is developed for VPP.A federated adversarial deep Q-network(DQN)based algorithm for robust optimization scheduling of VPP is proposed,where robust optimal strategies are achieved through iterative adversarial solution between two agents. Finally,to verify the effectiveness of the proposed algorithm and model,this paper constructs a simulation scenario of a power-communication-transportation coupling network based on an improved IEEE-33 node model.The simulation results show that,compared to two traditional algorithms,the proposed algorithm reduces the VPP economic cost by 15.13%and 24.33%,network losses by 14.83%and 19.76%,and node voltage fluctuations by 11.79%and 18.54%,respectively.Under adverse conditions,the proposed algorithm can still ensure the safety of node voltage,achieving optimal performance in extreme scenarios while maintaining robustness.The algorithm effectively enables efficient and robust control of the VPP while accounting for uncertainty factors in the power-communication-transportation coupling network.In future research,the author will consider the impact of real-time dynamic traffic flow changes on the optimization scheduling of VPP demand response.Additionally,to further enhance overall control benefits,the author will conduct a two-stage robust optimization study that takes into account the uncertainties in the three-network coupling,integrating market dynamics into the analysis.关键词
电力-通信-交通/虚拟电厂/不确定性/鲁棒优化调度/联邦对抗深度Q网络(DQN)Key words
Power-communication-transportation/virtual power plant/uncertainty/robust optimization scheduling/federated adversarial deep Q-network(DQN)分类
信息技术与安全科学引用本文复制引用
潘超,李梓铭,龚榆淋,叶宇鸿,孙中伟,周振宇..计及电力-通信-交通耦合网络不确定性的虚拟电厂鲁棒优化调度[J].电工技术学报,2025,40(15):4755-4769,15.基金项目
国家电网有限公司总部科技项目(52094021N010(5400-202199534A-0-5-ZN))和中国南方电网有限责任公司科技项目(1500002023030103JL00320)资助. （52094021N010(5400-202199534A-0-5-ZN）
