Preventive Maintenance Optimisation Using Evolutionary Hybrid Algorithm

This paper proposes a hybrid algorithm which allows us to optimize maintenance policy by adding the feature of choosing the action´s combination suitable to the best maintenance dates. The hybrid approach (called HGACS) combines an ant colony algorithm with a genetic algorithm. This combination is due to the optimized function which has two parts: the first one can be well improved by the ant colony algorithm whereas the second one can not be improved by this method because this sub-function evolves with time. We show that the hybrid algorithm can obtain good results faster than a classical approach especially when the problem depends upon a big number of variables. Hence the developed approach is more suitable for large scale optimization