Study On Robust Project Portfolio Selection And Scheduling Problem

Project management is about selection and support of multiple projects under the guidance of available source costs and enterprise strategic plan.There are many uncertain factors in the process of enterprise project management,such as uncer-tainty of cost expenditure and resource consumption.Usually,these uncertainties are very likely to cause the end of the entire research and development project.In order to make the uncertain cost and resources can be effectively configured and utilized,the decision-makers need to integrate the decentralized projects and en-sure that the selected projects meet the strategic objectives of the enterprise and maximize the profits of the enterprise.Therefore,this thesis studies the problem of multi project portfolio management from two aspects of project portfolio selection and project scheduling.This thesis has studied the R&D project portfolio selection and scheduling prob-lem based on the background of new drug research and development.Because the precise information about cost and resource consumption in the decision-making process of new drug research projects doesn't exist,and few reliable historical data can be obtained,the distributional information of uncertain parameters is usually unavailable.Therefore,the robust optimization method becomes the first choice to study the project management problem in this thesis.This thesis considers the cost and resource fluctuation at various stages of research and development.Based on box and polyhedron uncertainty sets,a robust project portfolio selection model is established to resist the uncertainty in the R&D process.According to the dual theory,the original model is transformed into an equivalent and tractable linear robust counterpart model,and the final objective is to obtain a robust solution with more stable performance.Further,in this thesis,a robust project scheduling opti-mization model is constructed and the solution method is discussed for multi project portfolio management with elastic startup time.Finally,by adjusting the different parametric values of conservative level to simulate various uncertain environments,we provide decision-makers with the optimal project portfolio selection policy and scheduling scheme under the multiple cases.The main work of this thesis include in the following five aspects:(i)The uncertain parameters are depicted based on the box and polyhedron uncertainty set;(ii)Under the uncertain cost and resources,a robust R&D project portfolio selection model is established,which provides the decision maker with a robust optimization framework for project selection problem;(iii)A robust R&D project scheduling model is established for multiple project sets with elastic startup time;(iv)The different parametric values of conserved level are adjusted to simulate different uncertain environments,so as to select and schedule projects under the current uncertain situation;(v)The numerical experiments are designed for two classes of project portfolio models and the results are obtained via optimization software.At last,the result analysis illustrates the modeling ideas and demonstrates the effectiveness of the proposed optimization methods.