A Modified Particle Swarm Optimization Based on Gird Method for Optimization Design of Spring of Slider Phone

• Main Authors: Shih-Yu Lin, 施育霖
• Other Authors: Hsin-Chuan Kuo
• Format: Others
• Language: zh-TW
• Published: 2011
• Online Access: http://ndltd.ncl.edu.tw/handle/25293537176964803279

碩士 === 國立臺灣海洋大學 === 系統工程暨造船學系 === 99 === In this research, a numerical population-based optimization algorithm is developed for optimization problems, the algorithm is established based on the particle swarm optimization incorporated with the grid method. To verify the proposed algorithm, we selected 11 benchmark functions, with variables interacted or non- interacted. Meantime, there are three different number of dimensions for each tested functions are considered, they are 30, 50 and 100 respectively. The results obtained by the proposed method are shown much improved than those from the particle swarm optimization. Further, two engineering optimization design problems, a complicated gear problem and a plane-truss structure problem are solved by the proposed method. The results determined demonstrated the proposed method provide the solutions the same as the best ones obtained by other different approaches. Finally, an optimization design of springs of a slider cell-phone is designed with 8 variables considered. The proposed method is applied on the mechanism to determine the combination of the eight variables to reduce the stress concentration as much as possible.