Proactive and reactive resource-constrained max-NPV project scheduling with random activity duration

This paper addresses the resource-constrained project problem in which activity durations are stochastic variables and the objective is to maximize the net present value of cash flow in the project. First, using the two classical time buffering methods, proactive scheduling optimization models are constructed to generate the robust schedules. Then, two reactive scheduling models with different objectives are proposed to adjust the baseline schedules when disruptions occur during their execution. For the NP-hardness of the studied problem, three heuristic algorithms, including tabu search (TS), variable neighbourhood search (VNS), and a mixed version of VNS and TS, are developed and compared with the multi-start iteration improvement algorithm through a computational experiment conducted on a randomly generated data set. In addition, based on the computational results obtained, the effects of several key parameters on the proactive and reactive scheduling results are analysed, and some managerial insights are obtained. The research in this paper has practical implications for contractors to improve the project profit in an uncertain environment.