Determination of dielectric properties of natural fiber reinforced polymer composite using adaptive neuro fuzzy inference system

Abstract

In this study, the dielectric properties of polylactic acid reinforced with natural fibers are objectified by using adaptive neuro fuzzy inference system (ANFIS). This method is considered to be economically feasible as compared to fabricating dielectric samples and measuring dielectric properties with apt operating conditions and advanced equipment. Also, past research has focused primarily on polymer dielectric properties which does not involve any natural fiber inclusion, such that this research will focus on producing ANFIS models for natural fibers which will then be used to calculate dielectric permittivity ( ), dielectric loss ( ) with respect to frequency dependencies. Furthermore, experimental results of dielectric constants and losses of polylactic acid (PLA) based composites will be analyzed from past research which used physical techniques to fabricate composites and will be compared with the results from ANFIS models. It was found that error computation of both the properties are found to be low and the percentage difference in output data is considered less. In addition, the ANFIS neural network had predicted most of the data that was trained, and the crisp output values were found to correlate with the experimental dielectric properties which makes this prediction model possible to use as an alternative approach to fabricating composites and testing.