风险负荷下研发网络相继失效建模与仿真

由于企业内外部环境不稳定和研发网络结构比较复杂，针对企业之间容易发生的相继失效问题进行研究。根据研发网络特点，引用加权无标度模型生成研发网络，从风险负荷视角出发，借鉴负载容量模型思想，从节点初始负荷、节点容量和失效节点负荷重分配3个方面构建研发网络相继失效模型，运用数值仿真方法揭示相关规律。结果发现，网络相继失效规模随着容许参数的增加而稳定不变，在达到某一临界值后便直线下降直至稳定，此时对应的临界值即为关键风险阈值；每个可调参数对应着唯一一个关键风险阈值，并随着可调参数的增大而减小；网络相继失效规模随着可调参数的增大而扩张，随着权重系数的增大而降低；平均度与相继失效规模不相关。

Modeling and Simulation of Cascading Failure in R&D Network under Risk Load

The interaction between the risk of enterprises due to the uncertainty of external environment and the complexity of R&D network's structure results in the occurrence of cascading failure. The paper cites the weighted scale free network model to generate a R&D network according to the characteristics of R&D network firstly. Then the paper establishes the R&D network cascading failure model from the definition of initial load, the load redistribution of the broken nodes and the definition of capacity. Finally, the paper uses the numerical simulation methods to reveal cascading failures mechanism. Simulation results show that the cascading failures scale of R&D network is stable invariability with the increase of tolerance parameter. When cascading failures scale obtains the critical value, it decreases linearly and eventually become stable. And the critical value is the key risk threshold. Each tunable parameter has the unique key risk threshold which has negative correlation with the tunable parameter. The cascading failures scale of R&D network has positive correlation with the tunable parameter and decreases with the increase of the weight. The average degree has no relationship with cascading failures scale of R&D network.