Multi-objective evolutionary decision support for design-supplier-manufacturing planning

Abstract

Modern enterprises utilize strategic and dynamic partnerships among designers, suppliers, contract manufacturers and customers to achieve efficiency and response to rapidly changing markets. There is a clear need for planning and decision support tools, and the availability of efficient and accurate multi-objective algorithms is critical to this field. This paper poses the distributed product development as a multi-objective assignment problem, and describes a new class of multi-objective optimization algorithms based on the principles of differential evolution. The multi-objective differential evolution (MODE) algorithm is shown to approach Pareto optimal solutions in a wide class of continuous and discrete problems, providing a practical tool for this domain. A case study of real product designs from the printed circuit board industry demonstrates the effectiveness of the discrete MODE algorithm and its potential value in a decision support system for complex product development.