Synthesis of the Base Models for Analysis of Finite Fluctuations Based on Filtering Algorithms

Analysis of finite fluctuations (AFF) is based on the assessment of the impact of changes in a set of factors on the change in response. It is an urgent task associated with the modeling of dynamic processes that arise in complex technological and organizational systems. One of the initial stages of solving this problem is the synthesis of basic models for AFF, which can be based on an analysis of the processes under consideration. A perspective direction in the field of analysis and modeling of dynamic processes is the addition of standard least-squares and least-modulus methods with a regularizing functional, which allows including various adaptive properties in the base models. The article discusses classic adaptive models and algorithms that allow to analyse the processes of various nature in structurally complex systems. In addition to a review of existing adaptive methods, the goal of the article is to propose a modification of the Boyd method based on the use of idempotent metrics. Considered models can be used as basic models for methods of analysis of finite fluctuations.