Tabu Search to Solve the Synchronized and Integrated Two-Level Lot Sizing and Scheduling Problem

Abstract

This paper proposes a tabu search approach to solve the Synchronized and Integrated Two-Level Lot Sizing and Scheduling Problem (SITLSP). It is a real-world problem, often found in soft drink companies, where the production process has two integrated levels with decisions concerning raw material storage and soft drink bottling. Lot sizing and scheduling of raw materials in tanks and products in bottling lines must be simultaneously determined. Real data provided by a soft drink company is used to make comparisons with a previous genetic algorithm. Computational results have demonstrated that tabu search outperformed genetic algorithm in all instances.