A Memetic Framework for Solving the Lot Sizing and Scheduling Problem in Soft Drink Plants

Abstract

This chapter presents a memetic framework for solving the Synchronized and Integrated Two-level Lot Sizing and Scheduling Problem (SITLSP). A set of algorithms from this framework is thoroughly evaluated. The SITLSP is a real-world problem typically found in soft drink plants, but its presence can also be seen in many other multi-level production processes. The SITLSP involves a two-level production process where lot sizing and scheduling decisions have to be made for raw material storage in tanks and soft drink bottling in various production lines. The work presented here extends a previously proposed memetic computing approach that combines a multi-population genetic algorithm with a threshold accepting heuristic. The novelty and its main contribution is the use of tabu search combined with the multi-population genetic algorithm as a method to solve the SITLSP. Two real-world problem sets, both provided by a leading market soft drink company, have been used for the computational experiments. The results show that the memetic algorithms proposed significantly outperform the previously reported solutions used for comparison.