Scheduling of resource constrained projects via genetic algorithm with multiple skills of resources in multi-project environment

This thesis presents a genetic algorithm with invasive weed optimization reproduction for the multi-skill resource constrained multi-project scheduling problem with global and local resources (MS-RCMPSP). The problem is constructed by considering skill and skill levels of the resources and geographical competencies of them. The precedence relationships among tasks belonging to multiple projects is also considering, when scheduling the projects. The proposed method is tested with 30 generated instances and these instances are categorized as small, medium, and large. The performance of the algorithm and the parameter selections are separated into small, medium, and large categories, and performance evaluations are made by comparing the standard genetic algorithm and invasive weed optimization. The effectiveness of the proposed algorithm has been validated by the computational results and the algorithm's performance in this thesis is compared with the standard genetic algorithm (GA) and invasive weed optimization (IWO).