Constrained multi-objective antenna design optimization using surrogates

A novel surrogate-based constrained multi-objective optimization algorithm for simulation-driven optimization is proposed. The evolutionary algorithms usually applied in antenna design optimization typically require a large number of objective function evaluations to converge. The efficient constrained multiobjective optimization algorithm described in this paper identifies Pareto-optimal solutions satisfying the required constraints using very few function evaluations. This leads to substantial savings in time and drastically reduces the time to market for expensive antenna design optimization problems. The efficiency of the approach is demonstrated on the design of an L1-band GPS antenna. The algorithm automatically optimizes the antenna geometry, parametrized by 5 design variables with performance constraints on three objectives. The results are compared with well-established multiobjective optimization evolutionary algorithms.