Preemptive Project Scheduling With Resources Of Multi-skill And Time-windows

Resource-Constrained Project Scheduling Problem(RCPSP) is the classic problem ofthe project management domain. All of the classic RCPSP assume that the tasks are notpreemptive, that is, once the task starts which needs to be succeeded till its end. But mostof the tasks are allowed to be splitted and executed more than once. As fast as theeconomic grows, the enterprise holds the strategy that the employees need to be with moreskills to increase the rate of the resource utilization and the enterprise’s competitive power.First, the resource worth of the preemptive project scheduling with multi-skillresource from the theory and real aspects has be analyzed. And a mathematics model ofthe multi-skill resource task split-able project arrangement problem has be built, accordingto a clearly judgment and description of the problem. To solve the model, the articledescribes the basic theory of the branch and bound algorithm, proposes the method whichutilizes the fulfill rate to optimize the branch and prefer process. At the same time, tojudge whether the resource constrained requirement is meet through the comparison of thetotal number of the resource when the resource has a single skill. Oppositely, while theresource has more skills, we use the method to judge whether the resource constrainedrequirement is ok: process the resource-skill matrix, code and decode the resourcearrangement scheme. Finally, we prove the branch and bound algorithm is effective andfeasible to solve the multi-skill resource task split-able project arrangement problemthrough the test of the Patterson instance set, the comparison of the target function andexecution time of the hybrid genetic algorithm, the particle swarm algorithm and thebranch and bound algorithm.Our research result, from the theory aspect, enriches the research content of theRCPSP; From the real aspect, we solve the problem when the enterprise face the projectscheduling problem in a dynamic condition, design a software to give the best optimizedscheduling suggestion when giving a specific project and resource, to help the projectmanager to schedule the project.