An Algorithm for Simulation-Optimization of Multi-Stage Flow Lines

Abstract: Mathematical programming representation has been recently used to describe the behaviour of discrete event systems as well as their formal properties. This paper proposes approximate mathematical programming models for the simulation-optimization of flow lines with finite buffer capacities. The approximation exploits the concept of time buffer, modeled as a constraint that put into a temporal relationship the completion times of two jobs in a sample path. The solution of the approximate model can be used to obtain an approximate integer solution and bounds on the variables of the exact model. The strength of bounds and approximate solution, in reducing the computation time needed to find a good feasible solution for the buffer allocation problem, and the performance of several meta-heuristics are tested using a large set of randomly generated problems.