Information measurement systems in the digital society

Abstract

Abstract The transition to the next technological order requires robotics and automation in the industry. Managing such process of production is impossible without scenario modeling, forecasting and diagnostic of possible emergencies based on measurement information received from various sensors. The models used must be adequate to the real state of the object, and not to its project, on the basis of which it was designed and manufactured. Such models are called twin models of physical objects. The objective of the research is to obtain the approaches to the creation of information-measuring systems capable of processing large volumes of measurement information for implementing dynamically changing twin models of physical objects, and to carry out continuous training and adaptation to changing conditions. The proposed method allows continuous extraction of knowledge from the measurement information, accumulation knowledge about the object during its life. It was shown that hybrid models with physical and mathematical substantiation and capability of learning based on measurement data are the possible solution.