A MODEL-BASED APPROACH TO MULTIDIMENSIONAL INTERPOLATION PROBLEMS

Abstract

In this article, the problem of forming a sufficient source of information for the optimal
management of a discretely defined multidimensional technological object was considered
as a problem of multidimensional interpolation. An algorithm for developing a mathematical
model of multidimensional interpolation using an empirical method is proposed. The idea of
achieving interpolation accuracy based on the requirements for the adequacy of regression models
was put forward. Based on the results obtained in the course of the research, it was possible
to calculate the interpolation values of the received signal with very high accuracy according to
several parameters of the multi-signal forecast models belonging to the same system. Methodologically,
the main focus of the research is on building an interpolation function for an object
with many parameters and a sufficiently large amount of information and its success.