Abstract:
Today, in a dynamic and competitive environment, establishing a maintenance strategy to significantly track and improve the performance measures of systems become a strategic necessity for most industries. As a static maintenance not adapting to the evolution of systems state does not meet the expectations of most industries. This thesis is a new contribution in the field of industrial maintenance optimization. The main objective of this work is to develop effective modeling approaches and methodologies to solve problems related to the optimization and evaluation of preventive maintenance of complex structure systems. To achieve this objective, on one hand, a new formulation of multi-objective maintenance optimization model is proposed to determine the best preventive maintenance planning, the proposed model includes two objectives: maintenance cost minimization and system availability maximization, the goal is to optimize these objectives simultaneously while respecting the spare parts budget constraint. The non-dominated sort genetic algorithm (NSGA-II) is used to select the optimal Pareto solutions. In the other hand, a dynamic methodology based on the dynamic Bayesian network is proposed for the evaluation and optimization of the performance measures of a multi-state system taking into account the effect of
different types of maintenance. Finally, these results show the effectiveness of this approach in the context of an evaluation process and in providing the opportunity to evaluate the impact of the choices made on the future measurement of systems performances. Through diagnostic analysis, intervention management and maintenance planning are managed optimally which allows to target the appropriate action plan and to improve maintenance management.
