Permissible tolerance parameters of logical nodes cannot be adjusted according to the state of system when analyzing cascading failures of the substation automatic system(SAS)by using the model of effective properties. To solve this problem

a complex network based on logical connections of SAS is designed. The concept of closeness centrality is proposed to measure the importance of a node in the information transfer

and the permissible tolerance parameters of logical nodes are dynamically adjusted. In this way

an improved model of effective properties for analyzing cascading failure of SAS is developed. On this basis

the system efficiency loss in T1-1 substation in different disturbance cases is analyzed. Experimental results show that the improved model is more suitable for analyzing cascading failures of SAS than the previous model. This work is supported by National Natural Science Foundation of China(No. 51277023)and Jilin Provincial Science and Technology Development Plan(No. 20140204071GX).