Bifurcation and recurrent analysis of memristive circuits

Bifurcation analysis is a powerful tool for studying nonlinear problems. However, its superiorities may short when considering acquired data sets instead of controllable chaos generator signals. In this case recurrent and statistical analysis may help to investigate the system. In this paper we compare several methods for estimation of number of scrolls in a memristive circuit model with the multi-scroll chaotic attractor. The techniques of histograms, kernel density estimation, moving average and recurrent analysis are considered. Two high-resolution dynamical maps are presented for the investigated circuit. The moving average method was found to be the most reliable and accurate method for constructing dynamical maps. The results of the work can be used for hidden attractor detection, multiwing systems research and studying other aspects of nonlinear dynamical systems.