Applications of neural networks to solving SMT scheduling problems a case study

Abstract

In today's surface mount technology (SMT) production systems, an SMT assembly line is designed to produce different types of products. The system setup cost is often sequence dependent and the job sequence has significant impact on the productivity of these systems. In this paper, a nonlinear mixed integer programming model was proposed to generate production sequence and to minimize total setup cost for a small size SMT production line in Western Canada. A neural-network approach was employed to solve this case problem. Problem features and detailed results are presented to illustrate the model and the solution approach.