Use of Particle Swarm in the maintenance plan optimization

Maintenance planning is formulated traditionally in terms of a multi-objective optimization problem where reliability, availability, maintainability and cost act as decision criteria and maintenance intervals act as decision variables. Recently, the study of the maintenance planning has been extended to consider in the optimization process the effect of human and material resources on reliability, availability, maintainability and cost models, in order to obtain a more realistic maintenance plan, as it has been observed that the non consideration of these resources may suppose large deviations in the unavailability and cost goals presumed at their optima. In this context, however, the optimization of the maintenance plan involves often a high number of decision variables. These types of problems have been solved in the past using meta-heuristics optimization techniques such as Genetic Algorithms. This paper proposes and demonstrates the good performance of another modern meta-heuristics technique called Particle Swarm Optimization. In this paper the maintenance plan optimization of a motor-driven pump group of a Nuclear Power Plant, considering as decision criteria the cost and reliability and as decision variables, the maintenance and test intervals, human resources and spare parts is performed applying the new tool.