About this Abstract |
Meeting |
MS&T21: Materials Science & Technology
|
Symposium
|
Advances in Dielectric Materials and Electronic Devices
|
Presentation Title |
Neural Networks, Graph Approach and Fractals Application on Electronics Parameters Determination and Prediction in Perovskite Ceramics |
Author(s) |
Vojislav V. Mitic, Ivana Radovic, Branislav M. Randjelovic, Srdjan Ribar, Cristine Serpa, Ivana Ilic, Aleksandar Stajcic, Vesna Paunovic, Branislav Vlahovic |
On-Site Speaker (Planned) |
Branislav M. Randjelovic |
Abstract Scope |
We modified BaTiO3 and applied fractal nature analysis to connect grain size, pore surface and particles’ Brownian motions with measured dielectric properties. The fractal analysis idea is inspired by self-similarity in nature biosystems, where the chaotic structures could be controlled by recognized geometry structure or just to have disorder controlled towards the order. Fractal regression curve can be used for the reconstruction of the bacteria and viruses (such as COVID-19) motion; trajectories of the virus motion could be obtained. Another approach to calculate properties at the grain boundary based on the bulk sample measurements was employment of artificial neural networks. Their biomimetic ability allows the usage of very simple computer operations solving complex mathematical problems. As the third method, we applied graph theory, developing 1D, 2D and 3D cases, calculating values at grain boundary via matrix count. At the end, the main goal is to establish fractal electronic neural networks. |