Chaotic regimes of a fractal nonlinear oscillator

In the paper, a fractal nonlinear oscillator was investigated with the aim of identifying its chaotic oscillatory regimes. The measure of chaos for a dynamic system is the maximum Lyapunov exponents. They are considered as a measure of the dispersal of several phase trajectories constructed under different initial conditions. To determine the maximum Lyapunov exponents, algorithms are used which are related either to the study of time series (Benettin's algorithm) or to the direct solution of an extended dynamical system (Wolff's algorithm). In this paper, the Wolf algorithm with the Gram-Schmidt orthogonalization procedure was used as the method for constructing Lyapunov's maximum exponents. This algorithm uses the solution of the extended initial dynamical system in conjunction with the variational equations, and the Gram-Schmidt orthogonalization procedure makes it possible to level out the component of the maximum Lyapunov exponent when computing vectors along phase trajectories. Further, the Wolf algorithm was used to construct the spectra of Lyapunov exponents as a function of the values of the control parameters of the initial dynamical system. It was shown in the paper that certain spectra of Lyapunov exponents contain sets of positive values, which confirms the presence of a chaotic regime, and this is also confirmed by phase trajectories. It was also found that the fractal nonlinear oscillator has not only oscillatory modes, but also rotations. These rotations can be chaotic and regular.