MULTIOBJECTIVE SUPPLIER SELECTION USING GENETIC ALGORITHM: A COMPARISON BETWEEN WEIGHTED SUM AND SPEA METHODS


Vladimir Rankovic, Zora Arsovski, Slavko Arsovski, Zoran Kalinic, Igor Milanovic, Dragana Rejman-Petrovic

Abstract: Supplier selection is one of the bas ic and most important activities of purchasing management. This acti vity often includes solving of multiobjective optimization problems with different and usually conflicting objectives. Modern supplier sele ction techniques involve novel multiobjective optimization algorithms based on computational application. In this paper supplier selection using genetic algorithm is presented. The authors used two different methods: weighted sum method and SPEA method. Weighted sum method belongs to category of Decision before Search methods. SPEA method is a memb er of Search before Decision group of methods. As criteria for selecti on optimization variance of quality and total costs are used. Results show that described me thodology can be applicable for the practical purposes. Fi nally, comparative analysis of two different methods, used in this research, is presented.

Keywords: Supplier selection, multiobjective optimization, genetic algorithms, strength Pareto evolutionar y algorithm, weighted sum method

DOI:

Recieved: 20.06.2011  Accepted: 24.11.2011  UDC: 65.012.32:681.513.5

Reads: 1320   

Download document




Impact factor

Card image

CiteScore

Card image

SciMagoJr

Card image