Many-objective Optimisation for an Integrated Supply Chain Management Problem

Abstract

AbstractDue to the complexity of the supply chain with multiple conflicting objectives requiring a search for a set of trade-off solutions, there has been a range of studies applying multi-objective methods. In recent years, there has been a growing interest in the area of many-objective (four or more objectives) optimisation which handles difficulties that multi-objective methods are not able to overcome. In this study, we explore formulation of Supply Chain Management (SCM) problem in terms of the possibility of having conflicting objectives. Non-dominated Sorting Genetic Algorithm-III (NSGA-III) is used as a many-objective algorithm. First, to make an effective search and to reach solutions with better quality, parameters of algorithm are tuned. After parameter tuning, we used NSGA-III at its best performance and tested it on twenty four synthetic and real-world problem instances considering three performance metrics, hypervolume, generational distance and inverted generational distance.