A Very Fast Simulated Re-Annealing Algorithm for the Automotive Trim Industry

摘要

This paper presents a novel effective optimization approach for the nesting problem in the automotive trim industry. A very fast simulated re-annealing algorithm (VFSRA) is developed for the placement of two-dimensional irregular sheets and stencils. Irregular profiles are represented by approximate polygons using geometrical processing of Weiler tuple. The search configuration space is considerably shrunk by using a technique of module slide. Compact nesting density is fulfilled by severe penalties of escaping and overlapping areas and normal penalty of unoccupied area defined in the cost function. The designed stencil motion strategy and annealing schedule are suitable. The comparison experimental results show that the VFSRA is effective for the trim parts nesting problem. In addition, it can be applied in other similar fields.