Abstract

This paper deals with discrete topology optimization and describes themodification of a single-objective algorithm into its multi-objectivecounterpart. The result is a significant increase in the optimization speed andquality of the resulting Pareto front as compared to conventionalstate-of-the-art automated inverse design techniques. This advancement ispossible thanks to a memetic algorithm combining a gradient-based search forlocal minima with heuristic optimization to maintain sufficient diversity. Thelocal algorithm is based on rank-1 perturbations; the global algorithm isNSGA-II. An important advancement is the adaptive weighting of objectivefunctions during optimization. The procedure is tested on four challengingexamples dealing with both physical and topological metrics and multi-objectivesettings. The results are compared with standard techniques, and the superbperformance of the proposed technique is reported. The implemented algorithmapplies to antenna inverse design problems and is an efficient data miner formachine learning tools.