SIMULATION CHAOTIC DYNAMICS OF RELATIVISTIC BACKWARD-WAVE OSCILLATOR WITH USING CHAOS THEORY AND QUANTUM NEURAL NETWORKS

Abstract

Nonlinear simulation and forecasting chaotic evolutionary dynamics of such complex system as relativistic backward-wave oscillator is treated using the new combined method, based on the chaos theory algorithms, concept of geometric attractors, and algorithms for quantum neural network simulation. It has been performed modelling the dynamics of multilayer photon echo neural network for the case of noisy input sequence. It has been performed analysis, modelling and processing the temporal dependence of the output amplitude for the backward-wave oscillator, described by system of the nonstationary nonlinear theory equations for the amplitude of electromagnetic field and motion of a beam. The data on the Lyapunov’s exponents, Kolmogorov entropy, correlation coefficient between the actual and neural networks prognostic rows, referred to a temporal dependence of the output signal amplitude of the nonrelativistic (relativistic) backward-wave oscillator are listed. The combining the advanced algorithms of the modern chaos theory, concept of a compact geometric attractors and one of the effective neural network algorithms, or, in a more general using an effective model of artificial intelligence etc, could provide very adequate and quantitatively correct description of temporal evolutionary dynamics of most complicated systems, in particular, in the field of modern ultrahigh-frequency electronics