Genetic algorithms for improving material utilization in manufacturing

Abstract

In modern production industries, computer aided systems have been improving the efficiency and convenience of the various stages of work. However, as the complexity of computerized production systems increases, various techniques are still necessary. The problem we addressed occurs in computer systems that automatically make manufacturing process plans in the metal grating manufacturing industry. In the system, the merging of tasks as a work unit is important to reduce the material loss. However, there is no guarantee that merging always reduces the material loss. So, operators must compare the material loss rates of diverse merging cases to find a near-optimal solution that provides a low material loss rate. In this paper, we apply genetic algorithms to search the near-optimal solution of a planning problem focused on the reduction of material loss. In order to reflect the domain dependent characteristics, we apply genetic algorithms in two levels related each other. © Springer-Verlag Berlin Heidelberg 2006.