On the investigation of plasma turbulence by the analysis of the spectra

The article describes the examples of the analysis of the experimental data spectra for identifying typical structures of processes forming plasma turbulence. The method is based on the original algorithm which is close to the one-sample bootstrap. The base model for description of the fine structure of stochastic processes is finite local-scale normal mixtures. For finding the statistical estimates (maximum likelihood estimates) well known EM algorithm is used. The efficiency of the proposed research technique is demonstrated for a number of spectra’s set obtained in different modes of low-frequency plasma turbulence.