Research Area:  Metaheuristic Computing

Abstract:

Tournamenting process plays an important role in the selection of best alternative in a knockout system. This concept is applied in the development of a hybrid algorithm based on differential evolution known as tournament differential evolution. In this work, a hybrid tournament differential evolution algorithm is applied on a two-warehouse inventory control problem of deteriorating item with the objective for determining the lot-size, maximum shortage level and cycle length of the concerned system. The corresponding inventory model is formulated with two separate warehouse facilities, partially backlogged shortages and all-unit discount. Here the demand of the product increases with time. The corresponding optimization problem of the developed inventory model is highly non-linear constrained optimization problem which cannot be solved analytically. In this context, to solve the optimization problem, the said hybrid algorithm is developed with the help of six different options of tournamenting procedure and differential evolution. Then, to examine the validity of the proposed model and also to test the performance of the hybrid algorithm, numerical experiments are performed by solving a numerical example and nonparametric statistical test. Finally, to investigate the effects of changes of different parameters, sensitivity analyses are carried out on the best found policy and a fruitful conclusion is drawn along with the future scope of research.

Keywords:  
tournamenting
knockout system
differential evolution
lot-size
maximum shortage level
backlogged shortages
unit discount
nonparametric statistical test

Author(s) Name:  Amalesh Kumar Manna, Md Akhtar, Ali Akbar Shaikh, Asoke Kumar Bhunia

Publisher name:  Elsevier

DOI:  10.1016/j.asoc.2021.107388

Volume Information:  Volume 107

Paper Link:   https://www.sciencedirect.com/science/article/abs/pii/S1568494621003112