Car sequencing for mixed-model assembly lines with consideration of changeover complexity

Implementing mass customization can inevitably lead to a large product variety, which makes the assembly process in mixed-model assembly lines (MMALs) very complex. In this paper, the concept of product variety induced changeover complexity, as one major source of uncertainty in mixed-model assembly, is proposed. Three types of changeover complexities measured using information entropy are presented. As the negative impact of changeover complexity on the performance of a MMAL can be reduced by selecting a suitable model sequence, a bi-objective car sequencing problem taking it into account is proposed. The problem is aimed at finding a model sequence with the minimum number of violating sequencing rules as a primary criterion, and the minimum level of total changeover complexity as a secondary criterion. A lexicographic approach based on ant colony optimization (ACO) is applied to solve the problem. Computational experiments show that both objectives can be effectively addressed using the presented ACO algorithm.