Comparison of evolutionary algorithms for LPDA antenna optimization

Pavlos I. Lazaridis, Emmanouil N. Tziris, Zaharias D. Zaharis, Thomas D. Xenos, John P. Cosmas, Philippe B. Gallion, Violeta Holmes, Ian A. Glover
2015 2015 1st URSI Atlantic Radio Science Conference (URSI AT-RASC)  
Key Points: • An LPDA antenna has been optimized by five evolutionary algorithms. • The best overall performance is exhibited by the IWO algorithm. • IWO produces the best fitness value but it also has the slowest convergence. Abstract A novel approach to broadband log-periodic antenna design is presented, where some of the most powerful evolutionary algorithms (EAs) are applied and compared for the optimal design of wire log-periodic dipole arrays (LPDA) using NEC (Numerical Electromagnetics
more » ... Electromagnetics Code). The target is to achieve an optimal antenna design with respect to maximum gain, gain flatness, Front to Rear ratio (F/R) and SWR (Standing Wave Ratio). The parameters of the LPDA optimized are the dipole lengths, the spacing between the dipoles, and the dipole wire diameters. The evolutionary algorithms compared are the: Differential Evolution (DE), Particle Swarm (PSO), Taguchi, Invasive Weed (IWO) and Adaptive Invasive Weed Optimization (ADIWO). Superior performance is achieved by the IWO (best results) and PSO (fast convergence) algorithms.
doi:10.1109/ursi-at-rasc.2015.7302885 fatcat:2l5fzsyouvefbl5r2k6j4joxbm