Multiobjective Dynamic-Guiding PSO for Optimizing Work Shift Schedules

Abstract

Work shift system is commonly used in construction projects to meet project deadlines. However, evening and night shifts raise the risk of adverse events and thus must be used to the minimum extent feasible. The three objectives of the work shift problem are to minimize project duration, project cost, and total evening and night shift work hours while effectively handling relevant scheduling constraints. This study proposes a new multiobjective approach that hybridizes dynamic guiding, chaotic search, and particle swarm optimization (PSO) functions, named multiobjective dynamic guiding chaotic search particle swarm optimization (MO-DCPSO). The approach can overcome the drawbacks of PSO in solving discrete domain problems and recruit more nondominated solutions kept in the archive. A real case was employed to verify the robustness and efficiency of the proposed approach. The result also indicated that MO-DCPSO is more fitting for solving practical project control issues.