Title:Multi-objective design of multilayer microwave dielectric filters using artificial bee colony algorithm

Abstract:Artificial bee colony algorithm (ABC) developed by inspiring the foraging phenomena of the natural honey bees is a simple and powerful metaheuristic optimization algorithm. The performance of single objective ABC performance has been well demonstrated by implemented to different design optimization problems from elec-trical and mechanical to civil engineering. Its efficacy is anticipated in the multi-objective design optimization of multilayer microwave dielectric filters (MMDFs) which is a challenging multi-objective problem falls into the context of electromagnetic (EM) design in electrical engineering. The MMDF is composed of superimposed multiple dielectric layers in order to pass or stop the EM wave at the desired frequency bands. An accurate dual-objective EM model of MMDF is constituted to be the total reflection coefficients for oblique incident wave angular range with transverse electric (TE) or transverse magnetic (TM) polarizations. The objective functions depend on the total reflection (TR) at the pass and stop band re-gions. Three types of the MDFs that are low pass (LP), high pass (HP) and band pass (BP) MMDFs are optimally designed across the constituted dual-objective EM model with a frequency dependent material database through using the multi-objective ABC (MO-ABC) strategy. The Pareto optimal solutions are refined within the possible solutions for synchronously minimizing the objective func-tions. The global optimal MMDF designs are selected from the Pare-to optimal solutions by providing the trade-off therein between the objective functions. The performance study regarding the frequen-cy characteristics of the designed MMDFs are well presented via numerical and graphical results. It is hence demonstrated that MO-ABC is even versatile and robust in the multi-objective design of MMDFs.