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

Abstract

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

Cite this work

No public copy is stored for this record, so only the metadata is cited.

Stored only when both papers are in the corpus — 25 citations exist upstream.

Papers that cite the same references — the relational fallback while embeddings are still backfilling.