Parameter Estimation in Design of 5.2 GHz Rectangular U Slot Microstrip Patch Antenna with ANFIS

A compact microstrip patch antenna (MPA) is essential for personal mobile and miniature communication. A simple approach based on the adaptive neuro-fuzzy inference system (ANFIS) is covered in this study for designing a rectangular U slot MPA. The significant benefit to the development of ANFIS model is that it entirely circumvents the antenna parameter calculation rather than running the IE3d repeatedly or other complex and time-consuming analysis techniques. The electromagnetic simulation program IE3D simulates 30 U slot MPAs with various physical dimensions for the ANFIS model’s training and testing, with twenty-three used for training and seven used for testing. The average absolute error was computed for tested datasets. At 5.2 GHz, the average absolute error for the tested datasets of the ANFIS model is found to be 0.57. The incredibly small error demonstrates the models’ accuracy. In CAD applications, the ANFIS model assists in ignoring repeated optimization time.