Vector evaluated particle swarm optimization (VEPSO): optimization of a radiometer array antenna
Dennis Gies, Yahya Rahmat‐Samii · 2004
Particle swarm optimization (PSO) has been receiving increasing interest from the engineering design community in the last couple of years. The successful application of genetic algorithms (GA) to a wide-range of engineering design problems has provided the impetus to investigate new and potentially better optimization techniques. At the time PSO was developed, significant research effort was focused on extending the evolutionary algorithm (EA) paradigms to multi-objective optimization problems. Many of the concepts developed for GAs have been modified to enable multi-objective optimization using PSO. The paper presents a study of the vector evaluated PSO (VEPSO), a multi-objective PSO paradigm, by applying it to a complex multi-objective electromagnetic design problem. We optimize the beam efficiency and half-power beamwidth (HPBW) of a 16-element array to be used to study soil moisture content from an airplane. This is accomplished by varying the element spacing and excitation amplitude of each element.