Antenna array optimization using surrogate-model aware evolutionary algorithm with local search

In the paper, antenna array optimization by means of an enhanced surrogate model-aware evolutionary search (SMAS) framework is presented. The enhancements aim at improving local optimization capabilities of SMAS. The local search engine, exploiting a surrogate assisted trust-region-based local search algorithm ORBIT is launched to fine-tune the solutions generated by SMAS to further decrease the optimization cost and maintain the global optimization ability of the algorithm. Two examples of linear array antenna problems verify the advantages of the method.