Neuro modeling of microwave waveguides with EBG structures

This work presents a new technique for analyzing microwave waveguide with electromagnetic bandgap (EBG) structures. Two types of waveguides, the rectangular waveguide and parallel plate waveguide, were modeled by neurocomputacional technique. The first waveguide uses the UC-EBG structure that consists of a metal pad and four connecting branches. This device is employed in the development of high-performance and compact circuit components for microwave and millimeter-wave frequencies. The second waveguide is formed between microstrip ground plane and metal plate with EBG elements on the top that works like one metal shield. This structure has capability of suppressing the propagation of unwanted parallel plate modes. The main objective of this work is the use of a novel ANN paradigm in modeling EBG structures. The proposed neural network model for microwave waveguides on EBG structures gave results in excellent agreement with measured results available in the literature. Besides, this technique is able to generalize for new valuables of input parameters and frequency ranges.