Oscilações periódicas e caóticas em sistemas diferenciais que modelam circuitos elétricos com memristores localmente ativos

Carregando...
Imagem de Miniatura

Data

2021-10-20

Autores

Sandoval, Liner Samer Albornoz

Título da Revista

ISSN da Revista

Título de Volume

Editor

Universidade Estadual Paulista (Unesp)

Resumo

O memristor (ou resistor com memória) é considerado o quarto elemento passivo fundamental de um circuito elétrico, em adição aos conhecidos resistores, capacitores e indutores. Neste trabalho, apresentamos um pequeno histórico sobre a descoberta deste quarto elemento, sua construção física e descrevemos algumas de suas principais propriedades, em termos das equações que o determinam e de exemplos. Em seguida, apresentamos o estudo de sistemas diferenciais em R2 e R3, que modelam circuitos elétricos envolvendo um memristor localmente ativo, controlado por corrente. Tais modelos foram propostos no artigo [12], que é a base do estudo aqui apresentado. O sistema diferencial planar, que modela um circuito envolvendo um memristor, um indutor e um resistor, possui cinco pontos de equilíbrio, duas selas, um nó estável, um nó instável e um foco, que pode ser estável ou instável, dependendo dos valores dos parâmetros. Usando a teoria das bifurcações, mostramos que tal sistema pode apresentar oscilações periódicas, provenientes de uma bifurcação do tipo Hopf, que ocorre no ponto de equilíbrio tipo foco, dando origem a um ciclo limite estável. Em adição a este resultado, obtido também em [12] por meio de propriedades do circuito elétrico, mostramos que o sistema planar pode apresentar uma órbita homoclínica, que delimita uma região no plano de fase no interior da qual o sistema apresenta oscilações periódicas, sendo que no seu exterior as soluções tendem para um ponto de equilíbrio assintoticamente estável. Desse modo, mostramos que as separatrizes dos pontos de sela desempenham um papel importante na dinâmica do modelo estudado. Vale destacar ainda que as oscilações periódicas ocorrem numa região em que o memristor é localmente ativo. No sistema diferencial definido em R3, que modela o circuito obtido adicionando-se um capacitor ao circuito planar descrito acima, além da ocorrência de oscilações periódicas, podem ocorrer também oscilações caóticas, dependendo dos valores dos parâmetros. Observamos numericamente que as oscilações periódicas existentes evoluem para caótica via uma cascata de duplicação de períodos, uma das conhecidas rotas para o caos, que ocorre ao variarmos um dos parâmetros do sistema. Novamente, as oscilações periódicas e caóticas estão relacionadas à atividade local do memristor. Ao longo do trabalho são apresentados gráficos de funções e de soluções dos sistemas diferenciais em R2 e R3, para um melhor entendimento do estudo, os quais são feitos utilizando os softwares MAPLE TM, MatLab e wxMaxima. Os estudos aqui apresentados contribuem para o entendimento da complexa dinâmica de circuitos elétricos envolvendo memristores como componentes ativos e, consequentemente, para a utilização desses componentes em áreas como a criptografia, construção de memórias não voláteis, redes neurais, inteligência artificial, dentre outras.
The memristor (or memory resistor) is considered the fourth fundamental passive element of an electrical circuit, in addition to the well-known resistors, capacitors and inductors. In this paper we present a short history of the discovery of this fourth element, its physical construction and describe some of its main properties, in terms of the equations that determine it and examples. We then present the study of differential systems in R2 and R3, which model electric circuits involving a locally-active, current-controlled memristor. Such models were proposed in the paper [12], which is the basis of the study presented here. The planar differential system, which models a circuit involving a memristor, an inductor, and a resistor, has five equilibrium points, two saddles, a stable node, an unstable node, and a focus, which can be stable or unstable depending on the parameter values. Using the theory of bifurcations, we show that such a system can exhibit periodic oscillations, arising from a Hopf-type bifurcation, which occurs at the focus-type equilibrium point, giving rise to a stable limit cycle. In addition to this result, already obtained in [12] by means of properties of the electric circuit, we show that the planar system can present a homoclinic orbit, which delimits a region in the phase plane inside which the system presents periodic oscillations, and in its exterior the solutions tend to an asymptotically stable equilibrium point. Thus, we show that the separatrices of the saddle points play an important role in the dynamics of the studied model. It is also worth noting that the periodic oscillations occur in a region where the memristor is locally active. In the differential system defined in R3, which models the circuit obtained by adding a capacitor to the planar circuit described above, in addition to the occurrence of periodic oscillations, chaotic oscillations can also occur, depending on the parameter values. We observe numerically that the existing periodic oscillations evolve to chaotic ones via a cascade of period doubling bifurcations, one of the known routes to chaos, which occurs when we vary one of the parameters of the system. Again, the periodic and chaotic oscillations are related to the region where the memristor is locally-active. Throughout the paper, graphs of functions and solutions of the differential systems in R2 and R3 are presented for a better understanding of the study, which are done using the software MAPLE TM, MatLab and wxMaxima. The studies presented here contribute to the understanding of the complex dynamics of electrical circuits involving memristors as active components and, consequently, to the use of these components in areas such as cryptography, construction of non-volatile memories, neural networks, artificial intelligence, among others.

Descrição

Palavras-chave

Memristor, Sistemas memristivos, Memristor localmente ativo, Oscilações periódicas, Comportamento caótico, Memristive systems, Locally active memristor, Periodic oscillations, Chaotic behavior

Como citar