A new operator that combines artificial immune systems and multi-agent systems for addressing flow shop scheduling problems

This paper introduces a novel mutation operator that integrates concepts from artificial immune systems and multi-agent systems to enhance systematic preventive maintenance policies within hybrid flow shop scheduling. The mutation operator is designed to make slight adjustments to the sequence, generating a new yet similar sequence. Its primary goal is to prevent convergence to a local optimum and to increase diversity within the population. Additionally, the mutation operator can be viewed as a straightforward form of local search. The proposed method draws inspiration from human body behavior, leading to the implementation of a problem-solving strategy focused on optimizing processing time through metaheuristic techniques. This approach combines multi-agent systems with elements inspired by the artificial immune system. The effectiveness of our method is consistently demonstrated throughout the paper. We apply this approach to three preventive maintenance policies aimed at maximizing availability and maintaining a minimum level of reliability in the production chain. The results indicate that our algorithm surpasses existing algorithms, taking into account the potential periodic unavailability of machines during production scheduling.