DYPE/PANI kompozit filmlerin sıcaklığa ve PANI katkı konsantrasyonuna bağlı olarak dielektrik parametrelerinin GRSA ile tahmini

Abstract

 GRNN model for the estimation of dielectric parameters of LDPE/PANI composites was investigated.  Dielectric parameters depending on temperature and PANI additive were estimated with the accuracy varying between 93% and 99%.  GRNN model can suggest optimum PANI additive level and working temperature for LDPE/PANI composites. Since estimate made with generalized regression neural networks (GRNN) provide both time and cost savings, application of GRNN to many disciplines including geotechnical engineering, stock market forecasts, weather predictions, etc. become very important in recent years. Particularly, the widespread use of GRNN in the field of physics and materials science will provide a significant advantage in time-consuming and costly processes. Figure A. The application of GRNN method in prediction of static dielectric constant. Purpose: This work is devoted to investigate the accuracy of estimation of dielectric parameters of polymer/polymer composite dielectric materials by GRNN. Theory and Methods: Low Density Polyethylene/Polyaniline (LDPE/PANI) films with the PANI contribution of 0.7 wt.%, 1 wt.% and 3 wt.% were produced by using hot press molding technique. The dielectric parameters (ε' and ε'') of the LDPE/PANI films were determined for 20 °C, 50 °C and, 80 °C by dielectric spectroscopy measurements realized by an impedance analyzer between 100 Hz-6.75 MHz frequency interval. The generalized regression neural networks (GRNN) method was applied to estimate the dielectric parameters of both experimentally prepared LDPE/PANI films for different temperatures (35 °C, 65 °C and 95 °C) and experimentally unproduced LDPE/PANI films consisting of PANI in the wt.% of 1.5 and 6 for the temperature interval of 20-95 °C. Results: As a result of comparison of experimental data and GRNN results, it was determined that GRNN is a reliable method for estimating dielectric parameters. Conclusion: It was concluded that the optimum PANI additive level for and the optimum working temperature for LDPE films can be determined for various technological applications without any experimental production by using GRNN model.

Application of Supervised Machine Learning Regression Algorithm to Prediction of Dielectric Properties of PPy/Kufeki Stone Composites for Energy Implementations

“…Considering ML algorithms, it is possible to estimate a given material's features including its structure [2,3], molecular properties [4,5], bandgaps [6,7], and dielectric properties [8]. Dielectric properties of a material can also be estimated by the artificial neural network method [9][10][11].…”