The Convolutional Approach for the Integration of Methods of Artificial Intelligence (AI) and Measurement Science (MS), Based on Bayesian Intelligent Technologies. The Concept of Bayesian Measurement Network. The Concept of IIIoT - Intelligent Industrial IoT

The article presents the concept of deep integration of artificial intelligence methods and measurement theory. This integration is implemented by convolving information flows in the form of data and knowledge at all levels of information acquisition and processing. At the same time, it improves the efficiency, stability, consistency, transparency, traceability and quality control of measurement and neural network solutions, the intellectualization of data transmission and storage systems (IIoT), and the construction of highly efficient cyber-physical systems. The paper considers the approach and methodological principles of creating a new type of neural networks, called Bayesian measurement networks. The concept and formalization of a new type of networks based on the regularizing Bayesian approach, Bayesian neurons implementing Bayesian convolution based on the regularizing Bayesian approach are given, an architectural scheme and metrological justification of solutions are proposed. The concept is given and a practical example of an intelligent industrial IoT-IIIoT is given. The expediency and experience of using Bayesian intelligent technologies for key areas of information systems development are noted.