Integration of Taguchi's Method and Particle Swarm Optimization for Electromagnetic Global Optimization

In this study, four optimization techniques, Taguchi's method [1], classical particle swarm optimization (PSO) [2], OA-PSO, and PSO-Taguchi's method, are investigated for performance comparison. To compare the performance of different approaches, test functions and a ten-element, unequally-spaced linear array are used as examples. Since random values are used in PSO-based approaches, 200 independent runs are conducted in each approach for statistical analysis. Numerical results and statistical analysis show that Taguchi's method can reach the global optimum quicker than other approaches; OA-PSO is more robust than classical PSO. (6 pages)