互联微电网韧性评估、调度与重构优化研究综述OA

The increasing frequency of extreme natural disasters and localized military conflicts poses unprecedented challenges to the stable operation of power systems.Networked microgrids(NMGs),which coordinate and interconnect multiple geographically proximate microgrids,have emerged as a promising solution to address these challenges.Compared to standalone microgrids,NMGs present significant advantages in increasing the penetration of distributed energy resources,improving power supply reliability,and enhancing overall system resilience.However,existing review studies on NMGs primarily focus on their structure and control strategies,with limited attention to the summary of resilience assessment,as well as mathematical modeling and optimization methods for resilient scheduling and reconfiguration.To address this gap,this paper provides a comprehensive review of NMGs,with particular emphasis on fundamental concepts and characteristics,resilience assessment indicators,and optimization methods for resilient scheduling and reconfiguration of NMGs. First,the paper defines NMGs and clarifies the relationships among NMGs,microgrid clusters,and multi-microgrids.It summarizes the main research modes of NMGs and explains the concepts of resilience assessment,resilient scheduling,and reconfiguration,along with their interconnections and distinctions.On this basis,the paper outlines the distinguishing features of resilient scheduling and reconfiguration in NMGs. The paper then reviews resilience assessment indicators for NMGs and organizes them according to their application forms and construction methods.In terms of application,resilience indicators are used as optimization objectives,operational constraints,or performance evaluation metrics.In terms of construction,they can be classified as either continuous or discrete types.Building on this classification,the analysis examines how different resilience indicators vary in identifying system vulnerabilities.Moreover,it points out that most existing assessment methods for NMGs still follow approaches developed for traditional distribution networks and mainly rely on load-side metrics,while the contribution of source-side units such as power electronic converters has not been sufficiently considered.Given the inherently multi-unit nature of NMGs,the paper accordingly proposes improvement directions from three perspectives,namely multi-unit,multi-dimensional,and multi-stage approaches. Subsequently,the paper focuses on the optimization strategies for resilient scheduling and reconfiguration of NMGs,providing a comprehensive summary of existing optimization objectives,constraints,models,and solution methods.The main optimization objectives in this context include minimization of overall system operating costs,reduction of the mismatch between generation and load demand,and maximization of load restoration.The key constraints mainly include power balance requirements,power trading rules,and limits on power exchange,among others.The paper reviews various optimization methods,including deterministic methods,uncertainty-based approaches such as stochastic,robust,and distributionally robust optimization,and deep reinforcement learning methods,which can be categorized into single-agent and multi-agent algorithms.On this basis,it analyzes the applicability,strengths,limitations,and future directions of these methods. Finally,the paper presents research prospects for NMGs.As research on NMGs is still at a developmental stage,future studies should focus on key areas including the construction of accurate uncertainty prediction models,the design of efficient distributed optimization methods,the enhancement of transient regulation and control in NMGs scheduling and reconfiguration,and the development of resilient networked multi-energy grids that leverage multi-energy collaboration to further improve NMGs resilience and operational efficiency.Additionally,the paper summarizes current challenges in the large-scale application of NMGs and provides relevant development recommendations.