A multi-objective optimisation algorithm for new distribution centre location

Abstract

Determining the location of a new distribution centre (DC) is a strategic decision that has critical implications on supply chain performance. This paper solves a model that formulates this decision as a nonlinear model with two objectives of minimising total supply chain and minimising inventory capacity on the two echelons of the supply chain. The paper presents and tests two multi-objective optimisation algorithms; non-dominated sorting particle swarm optimisation algorithm (NSPSO) and non-dominated sorting genetic algorithm (NSGA-II). For each algorithm, the paper tests three different settings for handling constraints. Both algorithms in the three settings outperformed published results. Analysis also showed that while NSPSO with its variations have competitive effectiveness over NSGA-II, they required longer CPU time.