Structural optimization under limited natural frequency constraints using comprehensive learning phasor particle swarm optimization

Abstract

In this research, a phasor particle swarm optimization with comprehensive learning strategy (CLPPSO) is proposed for the optimal design of dome-like truss structures under the limited frequency-constraints. The proposed scheme is a new variant of PSO techniques with the direct combination of both the phasor theory in mathematics and comprehensive learning strategy to the particle swarm optimization. In order to model particle control parameters, a phase angle incorporating the periodic sine and cosine functions is essentially applied through which only the previous best positions of all particles are used to update the exemplar particle’s velocity during the optimization process. This empowers the algorithm to keep the swarm's variability from eschewing premature convergence. To demonstrate the effectiveness and robustness of the proposed CLPPSO algorithm, three benchmarks of 120- bar, 600-bar and 1410-bar of dome truss structures are successfully tested, and the results are compared with those reported using different metaheuristic in the literature regarding their optimum solutions.