We consider the stochastic resource-constrained multi-project scheduling problem (SRCMPSP) where a number of projects with stochastic activity durations have to be scheduled such that the expected sum of the project makespans is minimized. In an experimental investigation we first generate a novel set of instances for the SRCMPSP subject to a well-defined test design with different problem parameters. We then perform a computational study employing a genetic algorithm with two different scheduling policy classes (resource- and activity-based) and study the performance of the latter with respect to the mean and the variance of the objective function.     «

We consider the stochastic resource-constrained multi-project scheduling problem (SRCMPSP) where a number of projects with stochastic activity durations have to be scheduled such that the expected sum of the project makespans is minimized. In an experimental investigation we first generate a novel set of instances for the SRCMPSP subject to a well-defined test design with different problem parameters. We then perform a computational study employing a genetic algorithm with two different scheduli...     »