Authors: Li Zhang; Jingui Lu; Ying Yu
Addresses: CAD Centre, Nanjing University of Technology, Nanjing 210009, Jiangsu Province, China; China Pharmaceutical University, Nanjing 210009, Jiangsu Province, China ' CAD Centre, Nanjing University of Technology, Nanjing 210009, Jiangsu Province, China ' China Pharmaceutical University, Nanjing 210009, Jiangsu Province, China
Abstract: This paper describes a successful adaptation of a discrete particle swarm optimisation (DPSO) to optimal design of bellows. EJMA standard and engineering practices-based mathematical model of unreinforced U-shaped bellows are established. An n-DPSO and scheme of constraint levels with huge value penalty are proposed to reproduce new particles and keep the best particles. Through benchmark problems, the obtained optimal results demonstrate that the approach clearly outperforms those from literatures. The bellows with diameters of 100 mm to 1,500 mm at 0.6 MPa to 1.6 MPa are designed by the approach and established mathematical model, the optimal solutions show that they are much better than those of bellows in-service and highly close to theoretical solutions. It is concluded that the established mathematical models are reasonable and the approach can be adopted in bellow design as it yields rational, reliable, time-saving and economically practical designs.
Keywords: discrete PSO; particle swarm optimisation; DPSO; huge value penalty; constraint levels; bellows design; mathematical modelling; optimal design; U-shaped bellows.
International Journal of Computing Science and Mathematics, 2014 Vol.5 No.1, pp.48 - 60
Received: 01 Jun 2013
Accepted: 20 Jun 2013
Published online: 30 Jun 2014 *