Research On Vulnerability Analysis Of Power Grid And Risk Control Strategy For Cascading Failure

With the steady advancement of construction of UHV grid and smart grid,the complexity of power system in grid structure and operation mode has been unprecedentedly increased,which introduces increasingly serious challenges to power system while solving the uneven distribution of energy and alleviating the imbalance between supply and demand.In recent years,large-scale blackouts occur frequently worldwide,which bring huge impacts on national economic development and normal life of residents.Cascading failure is generally regarded as the culprit of large-scale blackouts,therefore,exploring the evolution mechanism of cascading failure and conducting research on grid vulnerability analysis and risk control strategy have a great practical significance.The paper carries out study on modeling of cascading failure,vulnerability analysis of power grid and risk control strategy for cascading failure,the related work is summarized as follows:(1)Focusing on the tripping of overload lines and hidden failure,a cascading failure model based on AC power flow is established to simulate the propagation of overloaddominated cascading failure.Then the fast analysis method of cascading failure is studied where a cascading failure model based on DC power flow by simplifying the computation of power flow,and the probability distribution of load loss is estimated based on importance sampling.Finally,the advantages of the improved method in analyzing efficiency are verified through the simulation results.(2)Based on brief analysis of power grid vulnerability by complex network theory,the application of machine learning algorithm in key element identification is discussed.This paper proposes an identification method of critical lines based on FP-growth algorithm,which applies the FP-growth algorithm to mine the frequent items and to identify the critical lines.Then,an identification method of critical nodes based on modified PageRank algorithm is proposed,where the Google matrix of power grid is constructed by considering network power flow,node importance and other factors,and then the PageRank algorithm is used to sort the nodes to determine the critical nodes.Finally,the simulation and comparison results validate the effectiveness and superiority of the proposed method.(3)Form the perspective of dispatching operation and maintenance,the control strategy for dealing with cascading failure risk is studied in this paper.Firstly,a preventive control strategy considering the risk constraints of cascading failure is proposed and a solution scheme based on improved particle swarm optimization algorithm is designed.Case studies demonstrate that optimizing the operation mode of power system according to the proposed method can effectively reduce the risk of cascading failure.Then,a defense model for cascading failure based on preventive-emergency coordinated control is proposed.The model considers the complementary characteristics of preventive control and emergency control in which the system operation mode before the fault is optimized by preventive control and the propagation of cascading failure is cut off by the emergency control,so as to realize their complementary advantages in economy and safety control ability.Simulation results show that the coordinated control model can achieve the goal of economic optimization.