Author : Ojo O. Adedayo, Alli Kolapo and Ajibade Adedayo

Extraction of the dielectric properties of materials through open ended coaxial probe is a widely accepted microwave sensing technique and the prospect of application of soft computing techniques in this area is gaining significant attention. This paper therefore compares the performances of Multi-Adaptive Neuro-Fuzzy Inference System (ANFIS) and Artificial Neural Network (ANN) models for the complex permittivity of oil palm fruitlets which were obtained from laboratory microwave measurements in combination with the rational function model. The Multi-ANFIS consists of two ANFIS models which were designed, optimized and trained for the dielectric constant and the dielectric loss factor of the fruitlets, while the ANN was designed with two outputs to achieve the same aim from the same set of input data. The results show that while both soft computing techniques performed satisfactorily in modeling the dielectric properties of the samples under test, the unique ability of ANFIS to modify its human-friendly rules together with its efficiency even with reduced training data make it a readily preferable choice.

Keywords: ANFIS, Artificial Neural Network, dielectric constant, characterization, oil palm fruitlets.

Article published in International Journal of Current Engineering and Technology, Vol.5, No.3 (June-2015)