A “FAST” EXTREME FILTERING ALGORITHM

number of extrema is proposed that does not use decomposition into components with simultaneous smoothing. The possibility of estimating the parameters of time series in a sliding window and highlighting the modes in the analysis area directly by the extrema of the process is shown. The goal is to reduce the analysis time of large data sets and adapt the method to real-time mode. Materials and methods. The study was conducted in a Matlab environment using experimental data from a seismic detection system Results. It is shown that mode parameters can be estimated without decomposition based on the analysis of signal extremes. Conclusions. The proposed solutions can reduce the analysis time of rapidly variable processes, oriented to the use of real-time systems.