Research On The R&D Project Portfolio Management Using Complex Network Theory

The R&D project portfolio contains a large number of projects and the relationship between the projects.How to ensure that the selected projects are consistent with the corporate strategic goals and can effectively realize the corporate strategy?How to organize and manage these projects in the presence of numerous projects and limited resources,is the primary issue facing current decision makers.Therefore,based on the analysis of dependency relationship between projects,the paper uses complex network theory to manage project portfolio,and focuses on the following aspects:Firstly,using the complex network to select and priorize the projects in the portfolio.Aiming at the traditional project portfolio evaluation method that ignores the dependency relationship between projects,the paper uses complex network theory and PageRank algorithm to propose a new project priority ranking method.The results show that:considering the dominant relationship,as well as the diffusion relationship of technology and experience between R&D projects can better reflect the comprehensive influence of the projects in the portfolio.Also,the dominant and diffusion relationship have an important impact on the priority of projects.In practice,the project manager should comprehensively consider these relationships in the selection of projects.Also,improving the dominant and diffusion ability of the project in the portfolio can enhance its priority.Therefore,the project manager should identify the main factors affecting the change of project priority firstly,and then comprehensively analyze the priority of projects in the portfolio.Secondly,to solve the risk analysis problem in the project portfolio,the paper builds the two-layer "project portfolio-risk" network and proposes to use the restart random walk algorithm to analyze the project portfolio risk.The quantitative analysis method of portfolio risk proposed in the paper can more accurately assess the portfolio risk and rank the risk factors.The research results show that the dependency relationship between projects,the correspondence between the projects and the risk factors,and the initial dependencies between the risk factors will all affect the results of the project portfolio risk analysis,and the integration of the above three dependencies can accurately predict risk and its probability in the project portfolio,and priorize the risk factors.Thirdly,we analyze the risk propagation in the R&D project portfolio network.The paper uses the complex network and Bayesian network to build the risk propagation model to analyze the criticality of projects in the portfolio.The algorithm integrates the node attributes of the project and the edges embodied in the dependency relationship between projects,as well as the risk propagation to analyze the criticality of the projects in the potfolio.The results show that:the risk propagation makes the risk in one project will affect others in the portfolio,and even lead to the failure of the whole portfolio.Therefore,the risk propagation and its impact(including local and global impact)on project portfolio must be considered in analyzing project criticality.Fourthly,we analyze the knowledge diffusion in the R&D portfolio.To solve the problem of knowledge diffusion in the portfolio,the paper proposes a quantitative analysis method based on the SIR(Susceptible Infected Recovered)model,so as to measure the efficiency of knowledge diffusion more accurately.The algorithm analyzes the dynamic diffusion process of knowledge in the project portfolio network,and measures its effect on the knowledge level and schedule of the project portfolio.The results show that:the knowledge diffusion promotes the growth of the knowledge level,and also stimulates the willingness of knowledge diffusion between teams.Also,the knowledge diffusion reduces the total time and improves the success rate of the project portfolio.In all,the innovation of this study is as follows:1)using the complex network to solve the problem of project portfolio management.Based on the analysis of the whole process of the project portfolio management,the paper evaluates the problems encountered in the process of project portfolio management from four aspects:project portfolio selection,the analysis of project portfolio risk,risk propagation and knowledge diffusion;2)Using complex network and PageRank algorithms,the paper proposes the project priority ranking method for evaluating portfolios.The algorithm proposed in the paper can analyze the portfolio selection problem under the complex dependency relationship between projects;3)Based on the analysis of inter-project dependencies,the project portfolio risk analysis model,the risk propagation model,and the knowledge dynamic diffusion model in the R&D portfolio network are built.The research conclusions have important theoretical and practical significance for managers to clarify the dependency relationship between projects and control the whole process of project portfolio.