Partitioning For Reactive Voltage And Identification Of Critical Lines Based On Complex Network Theory

With the development of People’s demand for electricity,the size of power grid from the initial local small-scale is developed to today’s large-scaled interconnection,and the size of power grid is increased as well.The large-scaled interconnection within power grid will become a kind of trend.In the face of such a complex power grid,it is difficult to analyze power grid only by traditional reduction theory.Power grid can be conceptually described as a complex network,and it has the common characteristic of complex network,so complex network theory can be used to analyze it.Based on the topology of power grid,combined with the complex network theory,this paper will research the partitioning for reactive voltage and identification of critical lines.First,this paper summarizes a series of large-scale blackout accidents that occurred in recent years were counted,and introduces the basic knowledge and development of complex network theory.Then this paper focus on the current application of the complex network theory in power systemSecond,traditional partitioning algorithms for reactive voltage are often using clustering algorithm of power flow calculation.These algorithms have some problems,such as complex computation and unable to reflect the network structure of power grid.To figure out those problems,a new algorithm for voltage control partitioning is proposed based on spectral clustering of complex network.First,the model of voltage control partitioning is established using the imaginary part of nodal admittance matrix;second,the algorithm adopt spectral clustering to analyze the model to get the low dimension clustering samples;then power system is partitioned by the improving K-means clustering algorithm.To ensure the credibility of partitioning result,evaluation system with three indexes is established to calibrate the partitioning result,such as modularity,reactive power balance and reactive power reserve,etc.To improve the accuracy of pilot nodes selection,the evaluation index of degree centrality is introduced.The simulation result of IEEE 39-bus testing system shows that the new algorithm is effectiveness and superiority.Finally,in order to better identify the critical lines in power grid that may lead to cascading failures of power system,a new method based on comprehensive betweenness index and power transmission efficiency is proposed.Grounded on the power flow and distribution of power grid’s tidal current,this method has taken the power transmission margin of the line and post-breakdown transfer of power flow into full consideration,andquantified the role of critical lines in the power grid’s structure and motion,whose physical background is more accord with the practical situation of power system.Meanwhile,to validate whether the critical lines has a dominant role in the transmission of cascading failures,the power transmission efficiency index that has combined network transmission efficiency with directivity of power flow is put forward.The simulation result of IEEE-39 standard test system demonstrates that comprehensive betweenness index has a better performance in identifying critical lines in power grid.