Car Sequencing with respect to Regular Expressions and Linear Bounds

Abstract

In this paper we introduce a new model and a computational approach for sequencing assembly lines with two types of constraints: (i) patterns described by regular expressions and (ii) linear bounds on the number of certain products that may occur in pre-specified intervals. If we restrict the problem to the second type of constraints only we obtain a generalization of the familiar car sequencing problem, whereas constraints of type (i) may be useful to add extra structure. Constraints of both types may have priorities and can be violated, and a Pareto optimal solution is sought minimizing the violation of constraints in the given priority order. We describe a computational method based on mathematical programming and genetic algorithms for finding suboptimal solutions.