Allameh Tabataba'i UniversityIndustrial Management Studies2251-8029144220160922An integrated project portfolio selection and resource
investment problem to maximize net present value using
genetic algorithmAn integrated project portfolio selection and resource
investment problem to maximize net present value using
genetic algorithm61121570810.22054/jims.2016.5708FAHamidrezaShahabifardBehrouzAfshar-nadjafiJournal Article20141021In this paper, a mathematical model is proposed for project portfolio<br />selection and resource availability cost problem to scheduling activities in<br />order to maximize net present value of the selected projects preserving<br />precedence and resource constraints. Since the developed model belongs to<br />NP-hard problems list, so a genetic based meta-heuristic algorithm is<br />proposed to tackle the developed model. In the proposed algorithm beside<br />common operators of genetic algorithms such as crossover & mutation, some<br />intelligent operators are utilized for local search in computed resources and<br />shifting the activities with negative cash flows. The key parameters of the<br />algorithm are calibrated using Taguchi method to accelerate convergence of<br />the proposed algorithm. Then, the algorithm is used to solve 90 test<br />problems consisting 30 small-scale, 30 middle-scale and 30 large scale<br />problems to examine the algorithmâ€™s performance. It is observed that, in<br />small problems, the obtained solutions from the proposed genetic algorithm<br />have been comparably better than the local optimum solutions stemmed<br />from Lingo software. On the other hand, for the middle and large size<br />problems which there is no local optimum available within the limited CPU<br />time, robustness of the proposed algorithm is appropriateIn this paper, a mathematical model is proposed for project portfolio<br />selection and resource availability cost problem to scheduling activities in<br />order to maximize net present value of the selected projects preserving<br />precedence and resource constraints. Since the developed model belongs to<br />NP-hard problems list, so a genetic based meta-heuristic algorithm is<br />proposed to tackle the developed model. In the proposed algorithm beside<br />common operators of genetic algorithms such as crossover & mutation, some<br />intelligent operators are utilized for local search in computed resources and<br />shifting the activities with negative cash flows. The key parameters of the<br />algorithm are calibrated using Taguchi method to accelerate convergence of<br />the proposed algorithm. Then, the algorithm is used to solve 90 test<br />problems consisting 30 small-scale, 30 middle-scale and 30 large scale<br />problems to examine the algorithmâ€™s performance. It is observed that, in<br />small problems, the obtained solutions from the proposed genetic algorithm<br />have been comparably better than the local optimum solutions stemmed<br />from Lingo software. On the other hand, for the middle and large size<br />problems which there is no local optimum available within the limited CPU<br />time, robustness of the proposed algorithm is appropriatehttps://jims.atu.ac.ir/article_5708_138941f349eb14201e3220c677861c1d.pdf