Multi-Objective Fitness Functions with Non-Linear Switching for Antenna Optimizations

In a multi-objective optimization process, several goals are traditionally combined into single fitness function. such cases, the choice of objective function is critical, as it should accurately represent desired goals. Here, we introduce new class functions with non-linearity and switching behavior, also provide method for engineering. Notably, proposed versatile forms growth during optimization, systematic approach integrating expertise in antenna design process. The processes applied to demonstrate their enhanced performance. Our examples consider problems based on both analytical electromagnetic models full-wave simulation. Specifically, designs an end-fire array, pyramidal horn antenna, Yagi-Uda wideband patch antenna. results suggest that, minimum computation effort, non-linear produce better performing when compared linear summation-based function, e.g., 12% higher forward gain 9.3% lower side lobe level 23.38% directivity approximately 1.5 times bandwidth