Research On Robust Project Portfolio Based On CVaR

In an environment of increasing number of enterprises and increasingly fierce market competition,project portfolio selection is particularly important for the development of enterprises.Traditional project portfolio problems are mostly focused on maximization investment returns and the estimates of project returns is often subjective,leading to a lack of attention to risk and bias in return estimates.Therefore,the research on the project portfolio problem should comprehensively consider the influence of multiple factors to improve scientific and rationality.Aiming at the problem of grid project investment,a robust optimization model for the distribution of grid investment projects based on CVa R is established,using CVa R theory and robust optimization ideas.This model increases the emphasis on investment risk and weakens the subjective estimation of project returns.Due to the uncertainty of this model,the knowledge of duality theory is used to equivalent the model to a mixed integer programming problem including the SDP problem and the 01 problem.However,this mixed integer programming problem cannot be solved directly by the existing solvers.Thus,two algorithms is designed to solve the problem,the alternating optimization method and the conditional enumeration method.In order to verify the feasibility of the model and the algorithm,numerical calculations were performed through MATLAB.Furthermore,the comparison results showed that the calculation results of the two algorithms were similar and took less time when there were fewer options,when the number of selected items is large,the efficiency of the alternating optimization method is higher.Finally,the data experiments which use the actual data is performed to verify the applicability of the model and algorithms.Above results show that the optimization model proposed in this paper solves the drawbacks of traditional project portfolio problems to a certain extent.And the reasonable selection of the two designed algorithms is more convenient and efficient for solving practical problems.