Multi-objective optimisation for cost and carbon footprint over product life cycle文档.pdfVIP

MULTI-OBJECTIVE OPTIMISATION FOR COST AND CARBON FOOTPRINT OVER PRODUCT LIFE CYCLE ADA YUEN Abstract. This paper describes a graph based extension to the Strength Pareto Evolutionary Algorithm 2 (SPEA2). Two different approaches were tested, the joint and layer architectures, for minimising the cost and carbon footprint over the product life cycle. Cost and carbon footprint trade-off solutions produced from the joint and layer architectures are Pareto optimal sets on the Pareto-front. The two approaches were compared against the true Paerto-front in term of optimality, uniformity, extremity and run time to identify the best approach to the problem. The results show that the layer architecture can obtain better solutions than the joint architecture. Therefore the layer architecture with SPEA2G is considered to be the best approach for decision makers to solve multi-objective optimisation problems for cost and carbon footprint over a product life cycle. Keywords: Multi-objective optimisation; Multi-objective evolutionary algorithm; SPEA2; Product carbon footprint; Product life cycle 1. Introduction 1.1. Situation. In recent years the environmental impact of the product life cycle has become an important issue for businesses. Policy makers around the world are facing multi-disciplinary matters such as environmental pollution, resource consumption, regulation and global trade. Historically, maximising profits for corporate shareholders has been the ultimate goal of business. Businesses now face increasing resistance to their destruction of the environment. They need to