Research On Optimization Of Project Portfolio Risk Based On The Interval Programming

In order to keep the advantage of competitive, achieve the maximize benefits and reduce the management risk, the supervisor change the single project operation for the combination project operation. Then, the risk optimization and control of the project portfolio has been becoming the current hot area of research. But most of the research about portfolio risk measurement and optimization are focus on the financial field, and ignore the difference between the portfolio risk and capital portfolios risk on the risk measure and control optimization. In view of the problem, combing with the influence of interaction effect and the fuzzy uncertainty of portfolio risk information, this thesis study the optimization problem of the project portfolio risk with the information with interval numbers. The content is as follows:(1) The risk measure model of the single project is constructed, the risk of single project is evaluated through the entropy TOPSIS-AHP. Then, how to measure risk and how to evaluate single project with interval type are studied with PCA and Hausdorff distance and the corresponding numerical test are given by computer. The effectiveness of the method is proved by the experimental result.(2) The interval programming model of two projects portfolio risk based on the interaction effect are given. And using the interaction effect matrix theory and method, the two projects portfolio risk under resource, profit and technology factors are measured and optimized, moreover the model can adjust itself according to the different the type of risk. Then the model are solved through the Interval programming. At last, the corresponding program are coded by Matlab. According to the ratio of different type of portfolio risk, the comprehensive value of portfolio risk is given by solving the model, and the result shows that the projects portfolio risk has been reduced.(3) Considering the similarity between multi-project portfolio risk under interaction effect and complex network, the risk network of multi-project portfolio are established by the complex network theory. Then the network of multi-project portfolio risk is studied and optimized according to the characteristic of complex network. Furthermore it is solved by Matlab and Gephi and numerical experiment are given. The results show that the portfolio risk has been reduced and the important projects are saved. The effectiveness of the method is verified by the experimental result.By building the risk measurement and evaluation model of single project, the risk optimization model of two project portfolio and multi-project portfolio under interaction effect are established. This thesis research the optimization of project portfolio risk. The forementioned model and its code are given by the Matlab and Gephi or other mathematical software. While through analysing numerical example, the effectiveness of the methods are verified. Furthermore the efficiency and accuracy of calculate are improved. The theoretical foundation of methods for the manager to conduct the project risk measurement or evaluation and find the optimal portfolio risk are provided.