A Simulated Annealing Based Method for Sequencing Problem in Mixed Model Assembly Lines

The paper proposes a method to solve the mixed-model assembly line sequencing problem based on the Simulated Annealing Optimization algorithm. Achieving full line synchronization, by creating the appropriate model version sequence, becomes increasingly difficult at current levels of product complexity. The method of generating the candidate sequence by repeatedly swapping two random positions depending on the current temperature value was used. The search area is relatively large in the early phase of the algorithm. In addition, the conditions for resetting the temperature indicator if the local point candidate solutions are not improved have been added. It was also necessary to create a search objective function, taking into account specific aspects related to the mix-model sequencing problem. The proposed approach is based on binary coding of the input sequence and a suitably modified method of determining the boundaries of the search area. This increases the chance to avoid local optima trapping.