Optimization of the Preventive Maintenance for a Multi-component System Using Genetic Algorithm

The main goal of the manufacturers is to efficiently exploit their technological systems to improve their agility and productivity by using all available resources optimally in minimum time. For this, it is necessary to develop effective maintenance strategies to ensure the continuity of production and machine availability while minimizing the overall cost. Our work describes a policy adopted for the determination of the minimum cost of preventive maintenance of a multi-component system with respect to an availability constraint. For this, we have proposed a model that examines the situation of maintenance decision in which three actions, a minimal repair, a periodic overhaul and a complete renewal, the improvement of the system due to the maintenance action of a revision differs of the virtual age approach by considering a direct reduction of the failure rate. The genetic algorithm (GA) was used as a technique to optimize the cost function with respect to an availability constraint, i.e. to determine the optimum couples of periodicity of the preventive maintenance and the revision number. The results obtained considerably improved the preventive maintenance plan.