2025-03-16 Latest Published
By Rajendran Ramasamy
Maria Anto Bennet
Abbas Ali Farithkhan
Progress In Electromagnetics Research Letters, Vol. 125, 67-73, 2025
Abstract
In this research, the optimization of antenna parameters for the Vivaldi antenna, Inverted F antenna, and Probe Feed Microstrip Patch antenna was carried out using EOLRKC and the Sugeno Fuzzy Inference System (SFIS) machine learning techniques. The research explores numerical and conventional antenna design methods to understand the necessary concepts comprehensively. After a thorough analysis, an intelligent model for antenna selection recommends the best antenna based on various performance metrics evaluated with the Enhanced Logistic Regression Kernel Classifier. Additionally, the geometric properties of the antenna are discussed, and the SFIS is developed by integrating five primary learners to maximize the potential of each learner type. The EOLRK Classifier classifies antennas into three groups: Vivaldi, Inverted F, and Probe Feed Microstrip Patch, while SFIS determines the optimal parameters for antenna size. The accuracy of the EOLRK Classifier is assessed, while the performance of the Sugeno FIS is evaluated using MSE and MAPE. The proposed methodology achieves a MAPE below 4% and an accuracy exceeding 99%, demonstrating exceptional performance in parameter prediction and antenna classification. Implementing these methods has the potential to enhance innovative antenna design practices significantly.