Hybrid Real-Binary Particle Swarm Optimization (HPSO) in Engineering Electromagnetics

The applications of a hybrid real-binary particle swarm optimization (HPSO) algorithm in engineering electromagnetics are described. In HPSO, each candidate design is designated by a hybridized vector consisting of both real and binary variables. These variables are evolved in the optimization by following the velocity/position updating formulas of real-number PSO (RPSO) and binary PSO (BPSO), respectively. Both single- and multi-objective implementations of the algorithm are validated by functional testbeds. Simulation and measurement results of three examples, i.e, the design of a non-uniform antenna array, a multilayered planar radar absorbing material (RAM) and a dual-band handset antenna are presented, in order to illustrate the effectiveness of the algorithm in representative topology exploration and material selection problems.