Convergent Approach to Identification of Transient States of a Dynamic System

The article is devoted to the problem of identification of transient states of information sources (sensors, detectors, fractal biosignals, etc.) and optical media for information transmission. These states are often associated with self-organized criticality, which manifests itself near phase and structural transitions. However, the use of a variety of methods, models, patterns, parameters, indicators and criteria complicates the study of transition states. Stress factors cause temporal and spatial distortions of signals of different nature in the structure of interconnections which contains information about changes in the state of the information source. Therefore, transient states have common characteristic features, the versatility of which made it possible to implement a convergent approach to identifying transient states of information sources of different nature. The approach is based on the reconstruction of a topological 3D model from a measured scalar signal (time series, information flow). Orthogonal projections of the 3D model are the space-time signatures of the information source, which allow the analysis of the induced connections from three angles of view. Signature configurations consist of geometrically ordered sections that differ in steepness or curvature, as well as the interval between dynamic events. The characteristic features of transient states are the asymmetry of the antiphase components of the signatures and the imbalance in the powers of subsets of microstates. Visualization of 3D models and signatures of fractal signals of different nature simplifies the identification of transient states of the information source. Implementation of the approach will contribute to the effective selection of stress-resistant sensors (detectors, biosensors, etc.), as well as their monitoring.