Detecting communities of workforces for the multi-skill resource-constrained project scheduling problem: A dandelion solution approach

This paper proposes a new mixed-integer model for the multi-skill resource-constrained project scheduling problem (MSRCPSP). The interactions between workers are represented as undirected networks. Therefore, for each required skill, an undirected network is formed which shows the relations of human resources. In this paper, community detection in networks is used to find the most compatible working groups to perform project activities. In this respect, a greedy algorithm (GRA) is proposed to detect the most compatible communities of workers. The proposed greedy algorithm maximizes modularity as a well-known objective to find high-quality communities of workers. Besides, a new heuristic is developed to assign workers to activities based on the communities obtained by the GRA. The MSRCPSP is an NP-hard optimization problem with the objective of minimizing the makespan of the project. Therefore, a dandelion algorithm (DA), which is a meta-heuristic, is proposed to solve the problem. The dandelion algorithm is used to solve test problems of the iMOPSE</em> dataset. To validate the outputs of the proposed method, three other meta-heuristics including genetic algorithm (GA), harmony search (HS) algorithm, and differential evolution (DE) method are employed. The Taguchi method is hired to tune all algorithms. These algorithms are compared with each other in terms of several performance measures. The results show the superiority of the dandelion algorithm in terms of all performance measures.