Scenario-Based Solution Approach for Uncertain Resource Constrained Scheduling Problems

Many real-world decision problems involve uncertain parameters. The Resource Constrained Project Scheduling Problem (RCPSP) is one of those problems in which the activity durations are usually uncertain. Over the last decade, a good number of solution approaches have been developed to solve such problems, among them the population based algorithms received great attention. In the solution approaches, a large number of scenarios are usually evaluated which is computationally expensive. In this paper, as an attempt to reduce the computational time, we propose few alternative approaches and experiment them with an assumption that the uncertain parameters are random variables. For experimental study, these variables are generated using four different probability distributions. The proposed approaches are compared with the traditional scenario based approach by solving 10 well-known benchmark problems with 30 activities. The results revealed that it has advantages in terms of solution quality and computational time.