Preventive maintenance and replacement optimization on CNC machine using multiobjective evolutionary algorithms

In this paper, a multiobjective optimization model is formulated to determine the optimal preventive maintenance plans for a repairable multicomponent manufacturing system with increasing rate of occurrence of failure. The operational planning horizon is segmented into discrete and equally sized periods, and in each period, three possible maintenance actions (repair, replacement, or do nothing) have been considered for each component. The developed multiobjective nonlinear mixed-integer programming model is then utilized to find the Pareto-optimal preventive maintenance schedules for a computer numerical control (CNC) machine. In order to obtain the Pareto-optimal solutions, five multiobjective evolutionary algorithms are considered, and computational performance of the algorithms is evaluated using different metrics. It is found that three of these algorithms outperform the others in obtaining more and high-quality nondominated preventive maintenance and replacement schedules along with maintaining high level of diversity among generated solutions. Such a modeling approach and the comparison of algorithms would be useful for maintenance planners tasked with the problem of developing optimal maintenance plans for multicomponent manufacturing machines.