A hybrid solution approach under the resource constraints for maintenance scheduling problem

One of the major functions of a maintenance planning system is the ability of preparing the list of the Preventive Maintenance tasks to be performed in the next period together with indicating time, labour, materials, and tools feasibilities and priorities. So, maintenance scheduling can balance labour needed and can provide forecast for labour force that is going to be needed in the future. It is obviously possible to reach high degrees of performance with a good scheduled maintenance plan.Since scheduling is a NP-Hard problem, heuristic algorithms is preferred in the solutions of scheduling problems instead of analytical approaches. Genetic algorithms are quite efficient in scheduling problems of different sizes and gives successful results in obtaining feasible and better solutions. If it is not possible to perform the maintenance tasks in the time they are supposed to be conducted because of resource constraints (maintenance personnel) in scheduling problems, it is necessary to behave based on right criteria in deciding which tasks could be moved forward, which tasks could be postponed, or which tasks can not be delayed.In the system developed within this study, every preventive maintenance tasks in any given time space is represented as a gene. Lining up the genes in a row, chromosomes are composed that each one is a candidate for the solution. In the study, using the chromosomes in various arrangements considering the resource constraints, a genetic algorithm based solution technique is developed in order to assign tasks to previously defined dates taking account their priorities. Fuzzy Logic approach is preferred in defining priorities of preventive maintenance. The results of the developed algorithm in various sizes are compared with the test results of the traditional methods in various sizes.Keywords: Preventive Maintenance Scheduling, Genetic Algorithms, Fuzzy Logic