Comparison of SPEA2 and NSGA-II Applied to Automatic Inventory
Control System Using Hypervolume Indicator

Ewelina CHOŁODOWICZ, Przemysław ORŁOWSKI
West Pomeranian University of Technology Szczecin,
Sikorskiego 37, Szczecin, 70-313, Poland
cholodowicz.ewelina@gmail.com; przemyslaw.orlowski@zut.edu.pl

ABSTRACT: The optimization of multi-objective problems is an area of important research. The importance attained by this type of problems has allowed the development of multiple algorithms. To determine which multi-objective algorithm has the best performance with respect to the problem of goods flow in the inventory, in this article an experimental comparison between two of the main multi-objective evolutionary algorithms is conducted: Nondominated Sorting Genetic Algorithm II (NSGA-II) and Strength Pareto Evolutionary Algorithm 2 (SPEA2). The inventory model is optimized by taking into account two objectives: minimal cost of lost opportunities to make sales and minimal cost of used space in the inventory. The results obtained by both algorithms are compared and analysed based on hypervolume indicator that measures the volume of the dominated space.

KEYWORDS: Inventory control system; SPEA2; NSGA-II; multi-objective optimization; hypervolume.

Download PDF

CITE THIS PAPER AS:
Ewelina CHOŁODOWICZ, Przemysław ORŁOWSKI, Comparison of SPEA2 and NSGA-II Applied to Automatic Inventory Control System Using Hypervolume Indicator, Studies in Informatics and Control, ISSN 1220-1766, vol. 26(1), pp. 67-74, 2017. https://doi.org/10.24846/v26i1y201708