基于CA模型的研发网络级联失效建模及其对任务网络工期风险的影响研究

Modeling of R&D network’s cascading failure based on CA model and its impact on the duration risk of task network

对于研发项目而言,参研企业的级联失效很大程度上会影响研发任务的工期风险,这对研发项目工期风险管理具有重要启示。本文提出了基于研发项目的组织-任务关联网络模型,构建了基于CA模型的组织网络级联失效模型及其对任务网络工期风险的影响模型并进行数值仿真。结果表明,随着传染延迟时间的增长,稳态时正常节点以及负载节点的比例增长,失效节点的比例下降;随着网络投入成本μ以及控制参数ζ的增加,任务网络工期风险逐渐下降;度大袭击对任务网络工期风险的影响最大,其次是随机攻击和度小攻击;随着控制参数ζ的增大,随机攻击对任务网络工期风险的影响愈发明显,而度大攻击和度小攻击对任务网络工期风险的影响无显著变化。研究成果对研发项目工期风险管理具有一定的指导意义。

Faced with limited resources and complex customer needs for new R&D projects, many R&D project’s participant companies often take the form of cooperative research and development to achieve complementary resources and share risks. Therefore, R&D network has become the trend of the operation of R&D projects. Certainly, operating the R&D project in the form of the network will reduce cost and risk for participant companies. When some of the companies cannot operate well due to the deterioration of market conditions or their poor management, the failure event may affect other participating companies through the cooperative relationships among them, thus resulting in the cascading failure phenomenon among participant companies. As for the R&D project, cascading failure of participant companies will delay its implementation of the tasks, thus forming the task duration risk. Therefore, this study has practical implications because it considers cascading failure as a risk factor to study its effect on R&D task’s duration risk. In the first part, according to the basic characteristics of R&D projects and the development of the network, we built the organization-task related network from the perspective of system analysis. In the second part, considering the transmission delay influence on the organizations in the process of the cascading failure, we built the cascading failure model of the R&D project’s organization network based on the CA (Cellular Automation) model from the perspectives of infection rate, recovery rate, and failure rate. In the third part, we proposed a research model to understand the cascading failure’s impact on the duration risk of the task network, with the combination of the organization’s execution on the task and company’s failure extent. Then we did a numerical simulation of the model by changing the relevant parameters and using different attack strategies. The simulation results showed that: with the increase of the transmission delay time, the proportion of normal nodes and load-nodes increased, the failure nodes decreased. With the increase of the network investment cost, the duration risk of the task network decreased. With the increase of control parameter of transmition, the duration risk of the task network increased gradually, but the growth of the duration risk with corresponding value under certain level of network investment cost decreased. Among them, when we increased control parameter of transmition from 0.6 to 0.8, except for the condition when network investment cost is 0.1, the value of the duration risk of the task network were relatively close, and they got the same values under four different values of network investment cost, in this condition, the increase of the value of control parameter of transmition did not cause additional effects on duration risk of the task network. The impact of HD attack on the duration risk of the task network was the largest, followed by RD attack and LD attack, with the increase of control parameter of transmition, the impact of RD attack on the duration risk of the task network was becoming increasingly apparent, and the influence of HD attack and LD attack on the duration risk of the task network did not change significantly. In summary, this paper revealed the cascading failure rules of the R&D project’s organization network, as well as the influence of the cascading failure of the R&D project’s participate companies on the duration risk of the task network. The study provided a theoretical basis for the R&D project’s duration risk management under the background of network operation.