Maximização da eficiência energética de supercapacitores de grafeno usando RNA (Redes Neurais Artificiais))

Abstract

In this dissertation we present that supercapacitors can be prepared using simply computer simulations, provided that a series of procedures is followed. Electronics in general increasingly use energy storage prototypes in their configurations: capacitors and supercapacitors, so this topic is very relevant to research. The research itself aims to study the ANN (Artificial Neural Networks) method to obtain capacitors and supercapacitors, as well as calculate capacitances and test the main activation functions of a neural network: ReLu, Sigmoid, Softmax, Softplus, Tanh . However, this research has a main focus on (AI) Artificial Intelligence, as well as the achievement of maximizing the efficiency of graphene supercapacitors, using the RNA (Artificial Neural Networks - ANN) method. In addition, a characterization of the configuration of machine learning neural networks was developed, emphasizing strategies on how to obtain supercapacitors.