Research On Some Key Technologies Of Power Grid Operation Risk Assessment

With the deepening of power system reform,the integration of large-scale new energy,the development of long-distance large-capability electricity transmission and active distribution network,the modern power system has undergone significant changes.These changes have resulted in economic interests and social benefits,but also have brought many uncertainties at the same time,thus making safe and stable operation of power grid facing greater risks and challenges.Power grid operation risk assessment takes into account possibility and severity of grid failure,and thus can effectively analyze uncertainties and identifies power grid operation risks,so it is widely concerned.The key technologies of power grid operation risk assessment include component failure probability modeling,contingency screening,expected fault isolation,grid topology analysis,power flow analysis,load shedding,risk index system construction and calculation.In this paper,several key technologies are studied in depth,including contingency screening,expected fault isolation and grid topology analysis,and relevant methods are proposed,thus obviously improving the effecicency and practicability of power grid operation risk assessment form each link.Firstly,a multiple contingency screening method based on long tail distribution is proposed,avoiding the problem of multiple contingency combination explosion.This method can screen almost all severe contingencies in a fast way,overcoming the shortcomings of existing methods that cannot consider both screening rate and speed at the same time,and thus laying the foundation for accurate and quick identification of power grid operation risk.First,aimed at long tail distributed data,the data partition combination technology is proposed.Second,the branch overload contribution index is built,which is verified to be long tail distributed.Finally,the data partition combination technology is applied to the branch overload contribution index,and thus fast screening severe contingency.Secondly,an expected fault isolation method based on wiring mode is proposed,which accurately isolates fault component in a way that leading to the least non-fault component being isolated.This method takes account of the impact of station topology and wiring modes on fault isolation,filling the existing research gaps,and thus improving the accuracy of expected fault isolation.First,fault component is classified into independent component and component set according to grid wiring mode.Second,the component set analysis method is formed by analyzing network connectivity.Finally,fault isolation methods for different fault component are built.Thirdly,a power grid topology analysis strategy based on switch classification is proposed,achieving valid cooperation between station topology analysis and system topology analysis,thus improving practicability of power gird topology analysis,and avoding station topology analysis for unchanged station,thus improving analysis efficiency.First,switch is classified according to its affection on grid topology.Second,station topology analysis model and system topology analysis model are put forward by analyzing network connectivity.Finally,the coordination between station topology analysis and system topology analysis is studied,based on which the grid topology analysis strategy is constructed.Finally,network connectivity analysis is an important part of expected fault isolation and grid topology analysis,this paper puts forward a network connectivity analysis method based on node cluster growth.Compared with existing methods,the proposed method accelerates speed of network connectivity analysis.First,network topology model is introduced.Second,concept of node cluster growth is presented,and node cluster growth algorithm based on adjacent matrix is formed,which is used to analyze network connectivity.