PREDICTION OF RESONANCE FREQUENCY OF APERTURE COUPLED MICROSTRIP ANTENNAS BY ARTIFICIAL NEURAL NETWORK
Journal Title: International Journal of Engineering Sciences & Research Technology - Year 30, Vol 5, Issue 10
Abstract
In this study, the simulation model of Aperture-Coupled Micro-Strip Antenna (ACMA) by using Artificial Neural Network (ANN) is proposed. The developed model tries to predict the output resonance frequency of the ACMA according to the input physical parameters of the antenna. ACMA models were designed in High Frequency Structure Simulator (HFSS) software tool that could conduct three dimensional full-wave electromagnetic structure analysis based on Finite Element Method. Main objective is to simulate HFSS model via proposed learning model. Levenberg-Marquardt (LM) is utilized as a learning algorithm. 500 different ACMA models was designed in HFSS tool. Physical dimensions and output operating frequencies of the ACMA models were recorded in order to establish the dataset. Prediction performance of the proposed ANN simulation model was evaluated by 5-fold cross-validation scheme. Overall generalization error was calculated as 3.58 %. Experiments revealed that proposed simulation model operates at least ten thousand times faster than HFSS software. Due to its overwhelming running speed, it was concluded that proposed LM-ANN simulation model can be utilized as a preliminary search tool for optimizing the industrial ACMA models.
Authors and Affiliations
Musa Ataş
Keywords
Related Articles
- EP ID EP128672
- DOI 10.5281/zenodo.160855
- Views 42
- Downloads 0
A Two Tier Approach For Content Based Microscopic Image Retrieval System
Image retrieval is an active research area in image processing, pattern recognition, and computer vision. In this project, a multi-tiered CBIR system for microscopic images utilizing a reference database that cont...
Enhancing Digital Images Using Feed Forward Neural Network
A neural filtering technique is proposed to enhance the digital images when images are contaminated by impulse noise. This filter is obtained by combining nonlinear filter and feed forward neural network with back...
A KNOWLEDGE BASED APPROACH FOR CARRIAGE TAPE AND METAPHOR WITH NOMINATED ADMITTANCE PRIVILEGES
We observe the setback of scalable file system directories, forced by data-intensive applications require millions to billions of small files to be ingested in a lone register at charge of hundreds of thousands of...
AN ENHANCED UPGROWTH ALGORITHM FOR TEMPORAL HIGH UTILITY ITEM MINING
High utility item set mining from a transactional database helps to discover the items with high utility based on profit, cost and quantity. Even though many numbers of significant algorithms have been proposed in...
ENERGY MANAGEMENT IN ADHOC WIRELESS NETWORK
Wireless Adhoc Network (WAN) is infrastructure - less, self - organising, decentralised type of wireless network. Nodes acts as router. Battery management is an im portant issue in Wireless Adhoc Networ...