A multi-objective optimization approach to glass cutting problem

Abstract

2D glass cutting is an important problem for glass manufacturers. The objective of 2D glass cutting is to minimize the amount of waste when cutting a whole glass sheet into several pieces according to given cutting orders. In this research, another objective is introduced: Minimization of the defects on the glass pieces. The number of these defects defines the quality of the pieces, and each order delivered without defect or with minimum defect will increase the reputation of the manufacturer, as well as the competitiveness. Firstly, initial population strategy is used to generate first population for the genetic algorithm. Then an elitist evolutionary multi-objective optimization algorithm is applied for reaching Pareto front.