Rotational design space reduction for cost-efficient multi-objective antenna optimization

Cost-efficient multi-objective design of antenna structures is presented. Our approach is based on design space reduction algorithm using auxiliary single-objective optimization runs and coordinate system rotation. The initial set of Pareto-optimal solutions is obtained by optimizing a response surface approximation model established in the reduced space using coarse-discretization EM simulation data. The optimization engine is multi-objective evolutionary algorithm (MOEA). The designs obtained by MOEA are subsequently refined by means of response correction techniques. Our approach is demonstrated using a planar UWB dipole antenna.