To address the shortcomings of the existing cascading fault critical line identification methods in the specific quantization unit function and method complexity

combined with the complex network theoretical structure analysis method and the power grid cascading failure consequence approximate analysis method

a critical line identification method based on the improved LeaderRank algorithm is proposed. Firstly

the original LeaderRank algorithm is introduced

and the shortcomings of its direct application to identify critical lines of the power grid are pointed out. Then

aimed at the shortcomings of the original algorithm

a derived network based on the quantitative relationship of lines and a simplified accident chain model based on small probability event approximation are proposed. Finally

an improved LeaderRank algorithm is constructed based on the above model to identify critical lines of the power grid. The examples of IEEE39 system and actual power system show that the computational efficiency of the cascading failure model is significantly improved by using small probability event approximation. Moreover

combined with LeaderRank algorithm

it can more fully reflect the role of the unit in the system

including the system structure relationship

the degree of influence of the line unit

and the consequences of disconnection of the line unit. Thus the validity and effectiveness of the proposed method can be verified. The analysis method

which is based on the complex network theory to transform the grid topology into a derived network and is combined with a simplified accident chain model

not only reflects the severity of the consequences caused by the disconnection of the line

but also reflects the vulnerability of the line itself. The proposed method can comprehensively measure the criticality of a line in the power grid.