53 A Multistage Genetic Algorithm Approach for Solving Resource Constrained Multiple Project Scheduling Problems with Alternative Projects

In a multiple project environment, there also exists alternative ways for performing each project. This type of problem is called resource constrained multiple project scheduling problem with alternative projects (rc-mPSP∕aP). The rc-mPSP∕aP may be viewed as a network of projects which also involves project alternatives. In this research, we applied a monolithic approach which considers the alternative project selection and scheduling sub-problems together in a rc-mPSP∕aP where the activity times alter during the scheduling period. The monolithic approach formulates the problem as an exclusive scheduling problem and tries to solve the rc-mPSP∕aP. Particularly, a genetic algorithm (GA) approach was constructed in order to solving rc-mPSP∕aP efficiently. Since there exist two sub-problems in rc-mPSP∕aP, a multistage GA was created to reflect the sub-problems together in the exclusive problem. Finally, to evaluate the performance of the proposed monolithic multistage GA approach, various experiments were carried out on an example problem.