Application of Multi-objective Particle Swarm Optimization in Sheet Metal Forming

孙光永,李光耀,陈涛,张勇
DOI: https://doi.org/10.3901/jme.2009.05.153
2009-01-01
Journal of Mechanical Engineering
Abstract:Finite element simulation combined with optimization algorithm has been extensively researched in the sheet metal forming field to improve design quality and shorten design cycle. Conventional optimization techniques are mainly to transform multi-objective problem into a single-objective problem to solve. However, products are typically characterized by numerous quality characteristics (for example: rupture, wrinkling, insufficient stretching, etc.) and quality characteristics could conflict with each other. Therefore, it is of important significance to directly adopt a multi-objective optimization methodology to improve the formability of sheet metal. A parametric mould model is built by using an independently developed STLMesher software. On this basis, we obtain a set of non-dominated solutions which make wrinkle and rupture minimized simultaneously through combination of design of experiments, response surface models and multi-objective particle swarm optimization. In order to instruct engineering designers to quickly and effectively choose a set of solutions for best formability from the set of non-dominated solutions, the minimum distance solution method is proposed, with the selected solution realizing the optimization of wrinkle and rupture, hence enhancing the formbility of the sheet. The numeral example indicates that this method has higher precision and stronger engineering practicability.
What problem does this paper attempt to address?