The particle swarm optimization (PSO) method is becoming a popular optimizer within the mechanical design community because of its simplicity and ability to handle a wide variety of objective functions that characterize a proposed design. Typical examples arising in mechanical design are nonlinear objective functions with many constraints, which typically arise from the various design specifications. The method is particularly attractive to mechanical design because it can handle discontinuous functions that occur when the designer must choose from a discrete set of standard sizes. However, as in other optimizers, the method is susceptible to converging to a local rather than global minimum. In this paper, convergence criteria for the PSO method are investigated and an algorithm is proposed that gives the user a high degree of confidence in finding the global minimum. The proposed algorithm is tested against five benchmark optimization problems, and the results are used to develop specific guidelines for implementation.
On Global Convergence in Design Optimization Using the Particle Swarm Optimization Technique
Contributed by the Design Automation Committee of ASME for publication in the JOURNAL OF MECHANICAL DESIGN. Manuscript received January 3, 2016; final manuscript received May 19, 2016; published online June 20, 2016. Assoc. Editor: Christopher Mattson.
- Views Icon Views
- Share Icon Share
- Cite Icon Cite
- Search Site
Flocker, F. W., and Bravo, R. H. (June 20, 2016). "On Global Convergence in Design Optimization Using the Particle Swarm Optimization Technique." ASME. J. Mech. Des. August 2016; 138(8): 081402. https://doi.org/10.1115/1.4033727
Download citation file:
- Ris (Zotero)
- Reference Manager