Modeling and simulation of cascading failure on RD network based on different node states under attack strategies

摘要

Since cascading failure has negative and significant impact on R&D network, it is essential to research the characteristics of cascading failure for keeping R&D network safety. Firstly, the paper introduces the weighted scale-free network model to generate a R&D network. Then we develop overloaded function and restoration function to propose the cascading failure model of R&D network based on three states of nodes. Finally, the paper simulates the model. The results show that cascading failure can be stopped when the value of tolerance parameter exceeds the threshold value; the tolerance threshold value declines with the increase of tunable parameter, which indicates failure nodes with higher loads are easier to trigger cascading failure; the cascading failure size decreases with the increase of the value of restoration parameter; R&D network has strongest robustness under random attack, but is the most vulnerable to cascading failure under high-degree attack, which implies the importance of protecting high-degree nodes.