Coordinating Multiple Sources For Service Restoration In Distribution Systems After Large-Scale Outages

In recent years,large-scale outages due to extreme events happened worldwide.To deal with extreme events such as natural disasters,deliberate attacks,and cascading failures,constructing resilient grid has become a nation strategy for some countries.Improving the restoration capability of the distribution system is one important measure to enhance resilience.With power revolution and the development of the distribution system,multiple sources are integrated into distribution systems and the distribution system is equipped with monitoring and control devices,which makes it possible for the distribution system utilizing local sources to actively restore the critical loads during outage until the power supply from transmission system is restored.Utilizing multiple sources for service critical loads can shorten the outage duration and decease the outage losses,thus enhancing resilience.However,the types of local sources are various and the power and energy capacity of them is limited,so how to realize fast decision-making of restoration strategy to efficiently utilize multiple sources is challenging.To tackle the challenge,the research contents are as follows.(1)The restoration idea of multiple sources coordination and three mixed-integer convex optimization models for critical load restoration(CLR)decision-making problem by multiple sources coordination are proposed.First,how different types of sources can support the CLR is analyzed,and the idea of coordinating multiple sources for CLR is proposed,which can realize multiple sources coordination and complement with each other in space dimension.Then,the spanning tree constraints for radiality constraints are analyzed as the necessary but not sufficient conditions for radial topology,thus the single-community constraints are used to formulate radiality constraints.Based on different modellings of power flow constraint,three mixedinteger convex optimization models are built for basic CLR by multiple sources coordination problem and applicability of these models are discussed.Finally,the effectiveness of the model and the advantages of the restoration idea by coordinating multiples sources are validated by case studies.(2)Multi-period critical load restoration(MPCLR)by multiple sources coordination models and the MPCLR strategy based on rolling horizon are proposed.First,to tackle the challenges of long outage duration and limited generation resources,three mixedinteger convex optimization models for MPCLR are built,which can be solved by off-the-shelf solvers.The MPCLR idea can realize multiple sources coordination in spatiotemporal dimension.Then,to reduce the risk of load shedding and restoration failure by power uncertainties of intermittent sources,the MPCLR strategy based on rolling horizon is proposed,in which the predicted minimum power outputs of intermittent sources are only considered in decision making of operation plan.The effectiveness and advantages of MPCLR strategy are validated in case studies.(3)The MPCLR scheme considering mobile emergency resource(MER)dispatch is proposed.First,considering the influence of real-time traffic flow on MER dispatch,the optimization model of MER dispatch based on modified dynamic traffic assignment model is proposed.Then,the MPCLR model in chapter 2 is modified to consider the MER's arrival time,by solving which the dynamic restoration strategy can be obtained.Case studies validate the effectiveness and advantages of the proposed MPCLR scheme,which can make full use of MER to enhance the restoration effect.(4)Two-stage algorithm is proposed for solving the CLR and MPCLR models to satisfy online requirement.Based on heuristic idea and mathematical programming theory,the binary variables indicating line status are determined in the first stage to obtain a radial topology for restoration,and the load status binaries are decided in the second stage.For the first stage,two heuristics based on graph theory are proposed,i.e.,the iterative heuristic(IH)and the maximum spanning tree heuristic(MST).For the second stage,an iterative algorithm based on binary variable relaxation and cutting is proposed.Based on tests under large number of scenarios,it is shown that the proposed two-stage algorithm can obtain optimum and near-optimum with high efficiency,which is promising for online application.In summary,a multi-period on-line distribution system restoration decision-making method based on coordinating multiple sources is proposed,which can coordinate controllable sources,self-emergency power supply,intermittent sources,and mobile emergency resources to realize optimized allocation of limited generation resources in spatiotemporal dimension.The proposed method can enhance restoration capability,shorten outage duration,decrease outage losses,thus enhance distribution system resilience,which has profound significance to the safety of the society and the country.