Contextual Approach to Industrial Situation Recognition

Abstract

The article describes the development of multifaceted and efficient approaches to the context information analysis for synthesis of industrial situations context recognition algorithm in automated management systems within the enterprises. The probability theory method and method of statistical analysis, decision theory method, methods of algorithm and combination theory were used while researching. The research resulted in the development of new approaches to the context information analysis framework for pattern recognition which enables us to identify the procedure of contextual recognition for synthesis of working industrial situation recognition algorithm. A correspondence between the recognition error rate and the guaranteed recognition threshold, which can be used for setting up the automated context-based recognition systems, was analytically obtained during the research.