Granulation of multivariate time series in the problem of descriptive analysis of the state and behavior of complex objects

Being a source of implicit knowledge, multivariate time series (MTS) can act as models for the perception of objects in many applied areas. The article deals with the development of conceptual provisions for granular calculations of multivariate time series, on the basis of which a descriptive analysis technique is proposed that permits obtaining information granules about the state and behavior of the observed object expressed in textual form using protoforms. An application of granulation of a multivariate time series to the descriptive analysis of the development of the Russian economy is considered.