Algorithms for Detecting Changes in the Properties of Non-Stationary Processes in the Monitoring Mode

In this report, we have considered the problem of detecting and identifying the types of changes in non-stationary time series that lead to the need to rebuild the models describing the process. We compare two approaches to the construction of algorithms for detecting changes in the monitoring mode: fluctuation tests for detecting structural changes and sequential detection algorithms. The sequential algorithms have optimal properties in the sense of the criterion for the fastest detection.