New Wavelet-Based Approach Intended for the Analysis of Subtle Features of Complex Natural Signals

摘要

The present paper is devoted to the development of methods and algorithms intended for the anal ysis of complex natural signals (time series). Due to their variability, irregularity and complex structure the task of signal analysis and processing in the automatic mode is rather complicated. On the basis of contem porary methods of the analysis, processing, and recognition of complex data we have suggested a new approach, which allows us to automatically extract subtle features in complex natural signals of arbitrary structure. In addition, it becomes possible to identify components and characterize them in terms of a par ticular field. All the methods expounded in the following received approval from the Paratunka observatory (Paratunka, Kamchatka region, Far East Russia). The data were provided to the team of authors by the Uni versity of Cosmophysical Research and Radio Wave Propagation (Kamchatka region, Far East Russia).