Combining Random Sampling and Backward-Forward heuristics for Resource-Constrained Multi-Project Scheduling

In practice, organizations work in general on more than one project which share some or all the available resources. In this work hybrid heuristics that combine Random Sampling heuristics and Backward-Forward methods are developed for the Resource-Constrained Multi-Project Scheduling Problem (RCMPSP). This heuristic approach has been successfully applied to resource-constrained single project scheduling when the objective to optimize is makespan. In this work, the mean project delay and the multi-project duration increase criteria are considered. The heuristics proposed includes several parameters which effect is analyzed through a wide computational study. The best configuration of these parameters is reported when considering both criteria.