UDC 681.586
DOI: 10.36871/2618-9976.2021.12.003
Authors
Prokopchina S.V.
Doctor of Technical Sciences, Professor, Financial University under the Government of the Russian
Federation, Moscow, Russia
Abstract
The relevance of creating soft sensors for a wide range of practical
tasks is confirmed by examples of their use in various fields
of activity. The most important issue in this case is the development
of an effective methodology for their creation, focused
on the functioning of sensors in real practical conditions, most
often accompanied by significant information uncertainty.
The article notes that not enough attention is paid to the development
of the theory of the construction of electronic sensors,
as well as their metrological support by developers.
As a methodology for creating soft sensors for conditions of information
uncertainty, the article proposes a methodology of the
regularizing Bayesian approach and technologies based on it.
The effectiveness of the proposed approach is confirmed by the
practice of its application for the creation of soft sensors, examples
of which are given in the article.
Keywords
Soft Sensors
Regularizing Bayesian Approach
Metrology