A new method for examination of speaker identification is being offered. The base of the method lies in new theoretical statements and non-traditional views on the questions of formal transformation of voice signals and classification of voice objects. These views rest upon the concept of heuristic simulation. The novelty of a method is in the fact, that the identification is carried out on the base of the individual speech code, which is assigned to each voice code during the classification of voice objects adaptive search for informative tags. The digital code pattern of a personal voice code as a speed key of informative tags represents a set of elements with significant functions.