학술논문

Project Risk Response Decision Making Under Uncertain Project Interdependencies
Document Type
Periodical
Source
IEEE Transactions on Engineering Management IEEE Trans. Eng. Manage. Engineering Management, IEEE Transactions on. 71:7364-7378 2024
Subject
Engineering Profession
Optimization
Decision making
Risk management
Portfolios
Uncertainty
Research and development
Bayes methods
Polytopic fuzzy sets (PFSs)
portfolio risk
projects
response planning
Language
ISSN
0018-9391
1558-0040
Abstract
Project portfolio (PP) risk management entails making risk response decisions (RRDs) on single projects within PPs (SPPPs). However, the presence of uncertain project interdependencies (UPIs) increases the difficulty of RRDs. This study provides a decision model to make robust RRDs for SPPPs under UPIs, using a four-stage interval optimization model based on a two-tier risk-project network. A two-tier network is first constructed where the upper nodes denote the risks and the lower nodes denote the alternative projects. Next, a PP selection model is developed to obtain the probability of the projects being selected through simulation, in which the project interdependencies (PIs) are measured by the additional returns from the PP. Polytopic fuzzy linguistic term sets are formed to explore the strength of the relationships between the risks, and between the projects and risks. Third, the gravity model and the decision-making trial and evaluation laboratory are applied to prioritize the risks. Fourth, an interval optimization model under the uncertain response budget is constructed to yield the RRD outcome. The proposed model is validated through the RRDs for a Chinese construction firm. The results inform that, first, the mean utility of a decision maker (DM) considering UPIs fares better than those ignoring PIs and those considering PIs with certainty, second, the DM's mean utility considering PIs with certainty is still better than those ignoring PIs, and third, relying on the minimum budget threshold is more effective for reducing the uncertainty and improving the mean utility of a DM.