An efficient genetic algorithm for decentralized multi-project scheduling with resource transfers

<p style='text-indent:20px;'>This paper investigates the decentralized resource-constrained multi-project scheduling problem with transfer times (DRCMPSPTT) in which of global resources among different projects are assumed to be sequence-independent, while transfers local take no time within a project. First, two decision variables (<inline-formula><tex-math id="M1">\begin{document}$ {y_{ijg}} $\end{document}</tex-math></inline-formula> and <inline-formula><tex-math id="M2">\begin{document}$ {w_{ijg}} $\end{document}</tex-math></inline-formula>) adopted express transition state between projects. id="M3">\begin{document}$ (takes value 0 or 1) represents whether activity <i>i</i> resource <i>g</i> <i>j</i>; accordingly, transferred quantity is denoted as id="M4">\begin{document}$ $\end{document}</tex-math></inline-formula>. Then, we construct an integer linear model goal minimizing average project delay for DRCMPSPTT. Second, adaptive genetic algorithm (GA) developed solve To gain schedules DRCMPSPTT, traditional serial parallel generation schemes (SGSs) modified combine rules design multiple decoding schemes. Third, numerical experiments implemented analyse effects eight schemes, found that scheme comprising SGS maxRS rule can make GA work best; furthermore, effectiveness GA_maxRS (GA embedded best scheme) demonstrated by solving some instances sizes.</p>