Skobeltsyn Institute of Nuclear Physics
Lomonosov Moscow State University

This paper is about the multivariate statistical technique solving the problems of a certain kind: when we are unable to measure the interesting physical parameters directly, but we obtain some information about them and this information is multidimensional. The problems of such a kind appear in the CR physics if we use for the energy and another primaries characteristics determination such detectors, which allow to measure a large number of parameters depending on these characteristics.
In the network of the multivariate technique the algorithms of the primary energy and the primary CR spectral parameter  estimation were developed. They have been investigated on example of the concrete gaging equipment - the KLEM detector, NUCLEON project. The results of computer experiments modeling the multivariate technique work are presented. We have also performed a comparison between the suggested approach and the one-dimensional one. The comparison results demonstrate the advantages of the new method in both the accuracy and the universality as applied to practical realization.